business studies journal · 2017-04-04 · volume 2, number 2 issn 1944-656x h issn 1944-6578 o...

Volume 2, Number 2 ISSN 1944-656X HISSN 1944-6578 O

BUSINESS STUDIES JOURNAL

Gary SchneiderQuinnipiac University

The official journal of theAcademy for Business Studies,

an Affiliate of the Allied Academies

The Business Studies Journal is owned and published by the DreamCatchers Group,LLC. Editorial content is under the control of the Allied Academies, Inc., a non-profit association of scholars, whose purpose is to support and encourage researchand the sharing and exchange of ideas and insights throughout the world.

Authors provide the Academy with a publication permission agreement. Neither theAcademy, the Allied Academies, nor the DreamCatchers Group is responsible for thecontent of the individual manuscripts. Any omissions or errors are the soleresponsibility of the individual authors. The Editorial Board is responsible for theselection of manuscripts for publication from among those submitted forconsideration. The Editors accept final manuscripts in digital form and thePublishers make adjustments solely for the purposes of pagination and organization.

The Business Studies Journal is owned and published by the DreamCatchers Group,LLC, PO Box 1708, Arden, NC 28704 USA. Those interested in subscribing to theJournal, advertising in the Journal, or otherwise communicating with the Journal,should contact the Executive Director of the Allied Academies [email protected].

Copyright 2010 by the DreamCatchers Group, LLC, Arden, NC

iii

Business Studies Journal, Volume 2, Number 2, 2010


Gary Schneider, EditorQuinnipiac University

[email protected]

Board of Reviewers

Ismet AnitsalTennessee Tech [email protected]

M. Meral AnitsalTennessee Tech [email protected]

Santanu BorahUniversity of North [email protected]

Thomas BoxPittsburg State [email protected]

Steven V. CatesKaplan [email protected]

Susan ConnersPurdue University [email protected]

Carolyn GardnerKutztown [email protected]

Lewis HersheyFayetteville State [email protected]

Vivek Shankar NatarajanLamar [email protected]

Sanjay RajagopalWestern Carolina [email protected]

Ganesan RamaswamyKing Saud [email protected]

Durga Prasad SamontarayKing Saud University - [email protected]

David SmarshInternational Academy of [email protected]

Brian A. Vander ScheeAurora [email protected]

iv



TABLE OF CONTENTS

LETTER FROM THE EDITOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi

PERCEPTIONS OF ENCOUNTERS WITHDISRESPECTFUL STUDENTS: COMPARINGADMINISTRATORS’ AND BUSINESS FACULTY’SVIEWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Reginald L. Bell, Prairie View A&M UniversityShahedur Rahman, Prairie View A&M UniversityPeter W. Sutanto, Prairie View A&M UniversityAda Till, Prairie View A&M UniversityBettye Desselle, Prairie View A&M UniversityMunir Quddus, Prairie View A&M University

THE KEYS TO E-SERVICE RECOVERY:A FAST AND FAIR FIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Daisy Wang, Minot State UniversityChia-Jung Lin, Robert Morris UniversityKiattisak (Pong) Phongkusolchit, University of Tennessee at Martin

ARE S&P 500 INDEX MUTUAL FUNDSCREATED EQUALLY? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35John Cresson, Southeastern Louisiana University

2008-2009 TUITION AND FEES IN HIGHEREDUCATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Brian T. Kench, The University of TampaH. Scott Wallace, University of Wisconsin-Stevens PointKevin Neuman, University of Wisconsin – Stevens Point

v


CONVERGENCE OF E-TAILING ANDSOCIAL NETWORKING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Maria Luisa C Delayco, Hankuk University of Foreign StudiesBrandon Walcutt, Hankuk University of Foreign Studies

USE OF CREDIT CHECKS IN EMPLOYEE SELECTION: LEGAL AND POLICY ISSUES FOREMPLOYERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87Gerald E. Calvasina, Southern Utah UniversityRichard V. Calvasina, University of West FloridaEugene J. Calvasina, Southern University

DEVELOPING COLLABORATION SKILLS:A MIXED TEMPERAMENT APPROACH TOTEAMWORK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101David W. Nickels, University of North AlabamaJoan B. Parris, University of North AlabamaCarol H. Gossett, University of North AlabamaPaulette A. Alexander, University of North Alabama

PROACTIVE PERSONALITY:ORGANIZATION VS CAREER COMMITMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . 117Leila Messara, Lebanese American UniversityGrace K. Dagher, Lebanese American University

vi


LETTER FROM THE EDITOR

The Business Studies Journal is owned and published by the DreamCatchers Group, LLC. TheEditorial Board and the Editors are appointed by the Allied Academies, Inc., a non profit associationof scholars whose purpose is to encourage and support the advancement and exchange ofknowledge, understanding and teaching throughout the world. The BSJ is a principal vehicle forachieving the objectives of the organization.

The BSJ is a journal which allows for traditional as well as non-traditional and qualitative issues tobe explored. The journal follows the established policy of accepting no more than 25% of themanuscripts submitted for publication. All articles contained in this volume have been double blindrefereed.

It is our mission to foster a supportive, mentoring effort on the part of the referees which will resultin encouraging and supporting writers. We welcome different viewpoints because in thosedifferences we improve knowledge and understanding.

Information about the Allied Academies, the BSJ, and the other journals handled by the Academy,as well as calls for conferences, are published on our web site, www.alliedacademies.org, which isupdated regularly. Please visit our site and know that we welcome hearing from you at any time.

Gary Schneider, EditorQuinnipiac University

[email protected]

1


PERCEPTIONS OF ENCOUNTERS WITHDISRESPECTFUL STUDENTS: COMPARING

ADMINISTRATORS’ AND BUSINESS FACULTY’SVIEWS

Reginald L. Bell, Prairie View A&M UniversityShahedur Rahman, Prairie View A&M UniversityPeter W. Sutanto, Prairie View A&M University

Ada Till, Prairie View A&M UniversityBettye Desselle, Prairie View A&M UniversityMunir Quddus, Prairie View A&M University

ABSTRACT

In this study, we surveyed 400 faculty and administrators whose names appeared on acurrent directory of the Association for the Advancement of Collegiate Schools of Business-International (AACSB). We also surveyed 400 business faculty members from 14 randomly selectedHistorically Black Colleges and Universities (HBCU). Data were collected via Survey Monkey. Wecompared means of the responses and determined perceptions differed significantly between part-time and full-time faculty members and among other meaningful variables. Our findings suggestbusiness faculty and administrators who perceive they are encountering disrespectful studentbehavior should reconsider their views. We make a series of recommendations that highlight thesolutions to some of the more common problems of perceptions in relation to students’ disrespect.

INTRODUCTION

When students are disrespectful to the point effective teaching becomes impossible, nothingcan be worse for classroom teachers—and their students. In this study, we wanted to know iffaculty’s and administrators’ perceptions of their encounters with disrespectful students’ behaviorsvaried across employment status, gender, rank, nationality, and other meaningful variables. Whilethe literature is sparse on faculty perceptions of students’ disrespectful behaviors, even fewer studieshave compared empirically business faculty’s and administrators’ perceptions of their encounterswith disrespectful students’ behaviors directly with variables that might be significantly correlatedto their perceptions. It would certainly be a contribution to the management education literature to

2


know what aspects (gender, rank, part-time or full-time status, etc.) of the faculty’s workenvironment are significantly correlated with the differing dimensions of perceptions of encounterswith disrespect: our latent (hidden) variable.

We know student disrespect was evident as early as the nineteenth century when collegesexperienced student riots and disorderly behavior (Hessinger, 1999). Elite colleges and universitiessuch as Harvard, Yale, Johns Hopkins, Princeton, and the University of Pennsylvania suffered fromdisruptive student behavior (Brubacher & Rudy, 1997). College officials struggled to solve thedisorderly conduct as they moved from pre-colonial institutions that prided themselves on thepatriarchal authority of college faculty and administration to the Post-Revolution institutions wherestudents abhorred such authority.

Hessinger (1999) reported that during this period school authorities believed that the studentdisorderly conduct was caused by students’ loyalty to their peers over loyalty to their superiors.Officials at the University of Pennsylvania invoked a structure of college governance that includeda system of meritocracy by rewarding both good scholarship and good behavior. This system, knowntoday as grading, was originally developed at Yale in 1783 (Hessinger, 1999). The merit scalesadopted by many colleges in the early nineteenth century also included demerits, the loss of meritpoints for disorderly behavior (Hessinger, 1999). Historically Black Colleges and Universities(HBCU), which still primarily enroll African American students, also were encountering similarexamples of unruly conduct or what might be construed as “disrespectful behaviors.”

Up to the mid 20th century, more than 90% of African-American students enrolled in postsecondary programs attended HBCU. There were reports of significant student disrespect and unrestat Talladega College during the period of 1887-1914. The college, founded in 1867 by local AfricanAmericans and the American Missionary Association, was the site of a student rebellion in 1889,with students claiming that they were subjected to tyranny and outrage by the faculty. In 1895, theleader of a group of female protestors was suspended and sent home for disrespectful behavior. In1914, a group of female students went on strike against the disciplinary action taken by the faculty.By the 1920’s, African American students challenged white faculty and college administrators. Thisled to student unrest on black college campuses across the country (Jones, 1985).

The unrest on the HBCU campuses continued into the early 1960’s. Harrison (1972) notedthat student unrest on Black college campuses resulted in the resignation of presidents and otheradministrators, as well as faculty turnover. During this period, student grievances were related toseveral causes including disciplinary practices, social privileges, student personnel services, andfood services. By 1968, a good number of African American students were enrolled inpredominantly white colleges and universities, in part because of desegregation in higher education,and a dramatic increase in the number of African Americans enrolled in college (Kim & Conrad,2006; Willie & Cunnigen, 1981).

Nevertheless, HBCU continued to attract primarily African American students. In 2005, atthe approximately 100 HBCU, 83% of the students enrolled were African Americans, with a wide

3


variation in the ethnic population of the individual school. In 2005, student enrollment in all degree-granting institutions (approximately 4,300 two-and four-year institutions) was 17.5 million, whileenrollment at the HBCU was approximately 312,000. This was 1.8% of all college students and 14%of all African American college students (National Center for Education Statistics, 2007). Since theearly 1970’s, there has been no empirical research on disrespectful student behavior that exclusivelyapplied to HBCU.

As shown in the aforementioned studies, there is a need for this type of research as it isrelated to students at all institutions. There is a gap in the literature that needs to be filled becauseknowing what administrators and business faculty perceive about the disrespect from students theyencounter helps guide policy and classroom management activities in relation to these perceptions.There is a paucity of research on faculty perceptions of student disrespectful behavior at HBCU.These institutions were founded prior to 1964 for providing collegiate education to AfricanAmericans (Brown II and Davis, 2001). They have a unique history as compared to many of thecolleges that are members of the Association to Advance Collegiate Schools of BusinessInternational (AACSB). While less than a dozen HBCU are currently members of the AACSB, mostare not. Therefore, sampling merely from the list of AACSB members might result in anunderrepresentation of faculty from the HBCU whose views on disrespectful students’ behaviorsdeserve analysis too (Quddus, Bell, Bodie, Dyck, Rahman, Holloway, Desselle, & Till, 2009).

It is probably safe to assume that college business faculty members today encounter varioustypes of disruptive student behavior in their classrooms. The behavior may be overt such as talkingon a cell phone, or talking during class; or covert such as sleeping in class, arriving late, and eatingnoisily (Seidman, 2005; Meyers, 2003). Charles (1999) defined these disruptive behaviors asmisbehavior, while Hendrix (2007) referred to such behaviors as incivility, resistance, or reactivebehavior. The disruptive student may negatively affect the intellectual and academic progress ofother students. Such behavior can also impact the learning environment (Seidman, 2005). Facultymembers exposed to such behavior have experienced higher incidence of apathy, frustration, andteacher burnout (Evers, Tomic, & Brouwers, 2004).

The behavior of college students has changed because of changes in the ethnic and socialclass diversity, cultural norms, age, and lifestyle. Faculty members who possess pre-conceived ideasabout proper classroom etiquette will need to modify their expectations to reflect these changes(Boice, 1986:1993; Emerick, 1994; and Williams, 1994). For example, some faculty members mayassume that a student who wears a baseball hat indoors is inappropriately dressed, while otherfaculty members may not notice (Tom, 1998). In some cases, the manner in which a student dressesmay be viewed as an incivility, or insubordinate act, while the student’s behavior is based onignorance or what they perceive as current acceptable norms. Hendrix (2007) noted that one formof incivility involves the behavior of students of color toward professors of the same race at HBCUor at predominantly white universities.

4


Buttner (2004) studied business students’ opinions on respectful and disrespectful behaviorof business faculty. She found significant differences in expectations between female and malestudents in their opinions about how a faculty member conveys respect in the classroom. Buttner(2004) noted that the differences could be based on the differing views of female and male studentsof interactional justice, interpersonal relationships and interactions. Similarly, the perceptions ofmale and female faculty members may be affected by these same views.

In a recent survey of the members of the AACSB, the perceptions and actual disrespectfulencounters of faculty related to disrespectful student behavior revealed male and female businessprofessors’ perceptions and actual encounters with students’ disrespectful behavior were the same(Quddus, et al, 2009). This study also examined the affect on a business professor’s perceptions andactual encounters with disrespectful student behavior of the following variables: (1) a professor’srank (instructor, assistant professor, associate professor, full professor, or other); (2) contract status(tenured, tenure track, or non-tenure track); (3) employment status (full-time or part-time); (4)country of birth; (5) the Carnegie classification of the professor’s institution (community college,liberal arts, comprehensive, and research one); and (6) the professor’s ethnicity.

Quddus, et al (2009) found significant differences in the business professor’s perceptionswhen the professor’s rank and employment status were tested. Significant differences were alsonoted for professors teaching at various Carnegie classified institutions for actual disrespectfulencounters. They reported that student behavior that violated prudence (such as arriving to class late,or answering a cell phone), to be more disrespectful than behaviors associated with youthful self-expressions (wearing a baseball hat, or arguing about an assignment or grade). The researchers alsosuggested that a formal university-wide or college-wide code of conduct be developed: minorityfaculty was underrepresented in that study’s sample with 36 of 39 women reporting their race asWhite and 52 of 67 men reporting their race as White (Quddus, et al, 2009, p. 3).

This Study’s Purpose

This study’s purpose is based in part on the work of Quddus, et al. (2009). The purpose ofour research efforts is to compare the mean differences between perceptions of disrespect amongthe 400 administrators and faculty members surveyed in the AACSB directory and the 400 businessfaculty listed in the 14 HBCU randomly selected. This study will resolve the issue of samplingunderrepresentation of minority faculty in the Quddus, et al (2009) study.

The AACSB reports on its website (source: Business School Data Trends and 2009 List ofAccredited Schools) there were slightly more than 1,700 new doctorates in business awarded in2005-06. Not surprisingly, only 3.9 percent of the new doctorates in business went to Black, non-Hispanics in 2007-2008. The business faculty at AACSB accredited schools of business is woefullyunderrepresented by African American business scholars. Therefore, surveying a directory ofAACSB members only is less than adequate if the opinions of a diverse group of business faculty,

5


especially African Americans, are deemed important to faculty’s and administrator’s views ondisrespectful student behavior.

The main research questions more specifically asks: Are business professors’ demographiccharacteristics significantly related to their perceptions of their encounters with students’disrespectful classroom behavior? And, do faculty who are from AACSB accredited businessprograms differ from faculty whose affiliations are with HBCU organizations? To investigatefurther, the following hypotheses were written in order to test the research questions.

HYPOTHESES TESTING

The same Memory Encounters Scale (MES) developed and validated by Quddus, et al (2009)was used in this study. One-Way Analysis of Variance (ANOVA) was used to detect differences inthe MES with respect to (1) gender, (2) rank, (3) employment status, (4) contract status, (5) U.S.born, (6) code-of-conduct usage, and (7) ethnicity; we used MANCOVA to test mean differencesbetween (8) HBCU vis-à-vis AACSB affiliation, with employment status (full-time vs. part-time)used as the covariate. The null hypotheses were stated as follows:

Hypothesis 1: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES between maleand female professors.

Hypothesis 2: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES amongprofessors with different ranks.

Hypothesis 3: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES amongprofessors with different employment status.

Hypothesis 4: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES amongprofessors with different contract status.

Hypothesis 5: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES between US-born professors and foreign-born professors.

6


Hypothesis 6: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES betweenprofessors teaching at institutions with a mandatory code-of-conductand those without a mandatory code-of-conduct.

Hypothesis 7: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES amongprofessors with different ethnicity.

Hypothesis 8: There is no significant difference in perceptions of encounters withdisrespectful student behavior as measured by the MES betweenpersons listed in the AACSB directory and professors teaching atHBCU institutions when employment status (part-time and full-time)is used as a covariate.

METHODOLOGY

Survey, Sample and Descriptive Statistics

Ethical guidelines were followed and institutional permission was granted to collect data.A survey and a separate list of questions that pertain to assessment of demographic variables wereadministered electronically, via Survey Monkey. Four hundred names were randomly selected fromthe Association to Advance Collegiate Schools of Business-International (AACSB) list of 1,370.The sample consisted of professors and academic administrators representing collegiate schools ofbusiness throughout the world. The contacts were randomly selected, using systematic samplingfrom the AACSB population. Since it is well known that less than a dozen HBCU are accredited bythe AACSB, we wanted to assure ourselves we had a more diverse population of business facultyand administrators. This was a limitation in Quddus, et al (2009).

Therefore, a list of 400 faculty members names listed in the online directories of 14 randomlyselected Historically Black Colleges and Universities (HBCU) were surveyed. A random sample of14 of the 100 HBCU was selected. A sample of 400 HBCU email addresses was generated. Emailaddresses were verified prior to sending the survey. The survey responses were deemedrepresentative of the AACSB and HBCU population. A total of 111 out of 400 AACSB surveyswere useable since some respondents completed only the demographic portion of the survey and 36surveys could not be used for data analysis from the AACSB responses. Among the 400 HBCUemailed surveys, 61 responded and these were all useable. The 172 useable returns represent 21.5percent of those surveyed; 400 names from the AACSB directory and 400 names from 14 HBCU.

7


Analysis of the demographic data revealed 109 male professors and 61 female professors completedthe survey; two professors did not indicate their gender. The rank was represented by 13 instructors,20 assistant professors, 38 associate professors, 49 full professors, and 52 other faculty types.Among the respondents, there were 16 part-time and 155 full-time faculty, and 100 men and 53women on full-time status. There were 61 from HBCU and 111 from AACSB. Females and malesfrom AACSB represented 40 and 69 respectively, while HBCU females and males were 21 and 40respectively. The average part-time and full-time experience was 4.57 years and 15.2 yearsrespectively, with one person teaching for 45 years. There were 116 born in the United States, asopposed to 50 born elsewhere. Of those responding, 35 were from liberal arts colleges, 97 fromcomprehensive universities, and 39 were from research institutions. The contract status for malesand females combined was 43 non-tenure track (which comprised 25.4% of the respondents), 49tenure track (29%), and 77 were tenured (45.6%). And, finally, 110 White, 26 Black, 17 Hispanic,10 Asian and 4 other races of faculty members responded. The descriptive data collected from thesurvey responses is presented in Tables 1a, 1b, 1c, 1d, 1e, 1f, 1g, 1h and 1i.

Table 1a: Affiliation of Respondent

Frequency Percent Valid Percent Cumulative Percent

AACSB 111 64.5 64.5 64.5

HBCU 61 35.5 35.5 100.0

Total 172 100.0 100.0

Table 1b: Faculty Gender


Male 109 63.4 64.1 64.1

Female 61 35.5 35.9 100.0

Total 170 98.8 100.0

Missing 2 1.2

Total 172 100.0

8


Table 1c: Professorial Rank


Instructors 13 7.6 7.6 7.6

Assistant Professors 20 11.6 11.6 19.2

Associate Professors 38 22.1 22.1 69.8

Full Professors 49 28.5 28.5 47.7

Other Faculty types 52 30.2 30.2 100.0

Total 172 100.0 100.0

Table 1d: Employment Status

Frequency Percent Valid Percent Cumulative PercentPart-time 16 9.3 9.4 9.4Full-time 155 90.1 90.6 100.0Total 171 99.4 100.0 Missing 1 .6 Total 172 100.0

Table 1e: Contract Status


Non-Tenure Track 43 25.0 25.4 25.4

Tenure-Track 49 28.5 29.0 54.4

Tenured 77 44.8 45.6 100.0

Total 169 98.3 100.0

Missing 3 1.7

Total 172 100.0

Table 1f: United States Born


No 50 29.1 30.1 30.1

Yes 116 67.4 69.9 100.0

Total 166 96.5 100.0

Missing 6 3.5

Total 172 100.0

9


Table 1g: Carnegie Classification


Liberal Arts or Others 35 20.3 20.5 20.5

Comprehensive 97 56.4 56.7 77.2

Research Extensive 39 22.7 22.8 100.0

Total 171 99.4 100.0

Missing 1 .6

Total 172 100.0

Table 1h Ethnicity


Asian 10 5.8 6.0 6.0

Hispanic 17 9.9 10.2 16.2

Other 4 2.3 2.4 18.6

Black, non-Hispanic 26 15.1 15.6 34.1

White 110 64.0 65.9 100.0

Total 167 97.1 100.0

Missing 5 2.9

Total 172 100.0

Table 1i:Code of Conduct Mandatory


No 32 18.6 18.9 18.9

Yes 98 57.0 58.0 76.9

Not Sure 39 22.7 23.1 100.0

Total 169 98.3 100.0

Missing 3 1.7

Total 172 100.0

THE MEMORY ENCOUNTERS SCALE

As mentioned earlier, the same Memory Encounters Scale (MES) developed and validatedby Quddus, et al (2009) was used in this study. Quddus, et al (2009) showed the MES to have an

10


alpha reliability of .92. Since our study used a larger more diverse population we decided to run ourown alpha reliability test (Cronbach, 1951; 1984) on the MES. The same 16 items semanticdifferential, represented by anchors “1 = never” and “7 = all the time,” showed the MES alpha inour study to be .91, which exceeds the Nunnally (1978) criteria of 0.70; therefore, the MES appearsto be a “very good” measure of the construct: professors’ and administrators’ perceptions of theirown encounters with students’ disrespectful behaviors.

PRINCIPAL AXIS FACTOR ANALYSIS

Principal Axis Factoring, using Promax with Kaiser Normalization as a rotation method,revealed several variable loadings above .60 on the MES derived factors. The Un-rotated Sum ofSquared Loadings for the MES was 43.539, 11.433, and 7.360 respectively. In interpreting therotated factor pattern, an item was said to load on a given factor if the factor loading was .50 orgreater for that factor (Devellis, 1991; Hatcher, 1994; Kachigan, 1991) and was less than .50 for theothers. Statements S9, S11, S12 and S8 did not survive the factor rotation when the cutoff was .50or greater for a component to be said to load on a factor. The MES component saturation was veryhigh. Summaries of factor analysis results are shown in Table 2.

Table 2: Rotated Component Matrix for Three Derived Factors on the Memory Encounters Scale (MES)

MES Alpha Reliability = .91 on the 16 Survey Items

Three Factors explain 68% of the Variance in the MESF1:

DisagreeableF2:

UnresponsiveF3:

Attendance

S14: At least two students have stormed out of class in anger. .882 -.100 -.070

S15: At least two students have argued vehemently with me about anassignment or grade. .858 -.103 -.005

S16: At least two students have openly challenged my knowledge. .829 -.076 -.135

S4: At least two students in my class are argumentative with me andother students. .737 .034 .050

S10: At least two students treat me and other students discourteously(interrupting, horse laughing, bad mouthing, etc.). .628 .100 .145

S13: At least two students have answered cell phones while classwas in session. .561 .112 .041

S9: At least two students do not respond to my questions when I callon them. .447 .387 -.245

11

Table 2: Rotated Component Matrix for Three Derived Factors on the Memory Encounters Scale (MES)

MES Alpha Reliability = .91 on the 16 Survey Items

Three Factors explain 68% of the Variance in the MESF1:

DisagreeableF2:

UnresponsiveF3:

Attendance


S7: At least two students are not attentive during the lecture(sleeping, reading unrelated material, talking, browsing the Web,email of text messaging, etc.).

-.012 .818 .022

S6: At least two students have come to class unprepared (not readingrequired materials, having school supplies, and submitting requiredwork on time).

-.190 .782 .089

S5: At least two students choose not to participate in mandatoryclass assignments or activities. .081 .745 -.153

S11: At least two students talk to other students during my lectures. .250 .365 .209

S2: At least two students have missed several days of class withoutprior notice. -.050 .049 .819

S1: At least two students have repeatedly come to class late withoutprior notice. -.074 .177 .760

S3: At least two students have walked out of the class withoutpermission. .190 .089 .563

S12: At least two students wear improper dress in my class (baseballhat, t-shirt with profanity, loud-jingling jewelry, etc.). .388 -.019 .403

S8: At least two students address me by my first name in class. .186 .225 -.366

Extraction Method: Principal Axis Factoring. Rotation Method: Promax with Kaiser Normalization. A rotationconverged in 5 iterations.

Factor Rotation Sums of Squared Loadings(a)

Total Variance Explained

Total

Disagreeable 5.650

Unresponsive 5.030

Attendance 4.492

12


KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .896

Bartlett's Test of Sphericity Approx. Chi-Square 1308.453

df 120

Sig. .000

Guadagnoli and Velicer (1988) dealt directly with the relationship of sample size to thestability of component patterns. The authors used a Monte Carlo procedure to vary sample size,number of variables, number of components, and component saturation in order to examinesystematically the condition under which a sample component pattern becomes stable relative tothe population. They determined that factor loadings of .60 or higher with at least four variableloadings per factor permit stability of the factor pattern regardless of sample size. Furthermore,using SPSS 15.0 for this study, the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was.896, far exceeding the most conservative cutoff of .60.

RESULTS

We were unable to reject Hypothesis 1: There is no significant difference in perceptions ofdisrespectful student behavior as measured by the MES between male and female professors.ANOVA results are shown in Table 3.

Table 3: Gender

Sum of Squares df Mean Square F Sig.

Disagreeable Between Groups 2.579 1 2.579 1.905 .169 Within Groups 211.184 156 1.354 Total 213.763 157

Unresponsive Between Groups .631 1 .631 .443 .507 Within Groups 222.181 156 1.424 Total 222.812 157

Attendance Between Groups 1.124 1 1.124 .860 .355 Within Groups 203.930 156 1.307 Total 205.054 157

13


We were unable to reject Hypothesis 2: There is no significant difference in perceptions ofdisrespectful student behavior as measured by the MES among professors with different ranks.ANOVA results are shown in Table 4.

Table 4: Rank


Disagreeable Between Groups 3.695 4 .924 .673 .612 Within Groups 210.068 153 1.373 Total 213.763 157

Unresponsive Between Groups 3.591 4 .898 .627 .644 Within Groups 219.221 153 1.433 Total 222.812 157

Attendance Between Groups 5.437 4 1.359 1.042 .388 Within Groups 199.617 153 1.305 Total 205.054 157

We rejected Hypothesis 3: There is a significant difference in perceptions of encounterswith disrespectful student behavior as measured by the MES between professors with employmentstatus as full-time and part-time. The means for the 15 part-time and the 142 full-time employeeswere -.86 and .10 respectively. The significant difference was p= .003. ANOVA results are shownin Table 5. A full-time faculty member is more likely than a part-time faculty member to perceivean encounter with an unresponsive student as disrespectful.

Table 5: Employment Status


Disagreeable Between Groups .175 1 .175 .127 .722 Within Groups 212.719 155 1.372 Total 212.893 156

Unresponsive Between Groups 12.458 1 12.458 9.291 .003** Within Groups 207.834 155 1.341 Total 220.292 156

Attendance Between Groups .589 1 .589 .455 .501 Within Groups 200.662 155 1.295 Total 201.251 156

**Denotes significant difference at p< .01.

14


We were unable to reject Hypothesis 4: There is no significant difference in perceptions ofencounters with disrespectful student behavior as measured by the MES among professors withdifferent contract status. ANOVA results are shown in Table 6.

Table 6: Contract




Attendance Between Groups .938 2 .469 .357 .700 Within Groups 199.541 152 1.313 Total 200.479 154

We were unable to reject Hypothesis 5: There is no significant difference in perceptions ofdisrespectful student behavior as measured by the MES between US-born professors and foreign-born professors. ANOVA results are shown in Table 7.

Table 7: U.S. Born


Disagreeable Between Groups 1.001 1 1.001 .716 .399 Within Groups 211.317 151 1.399 Total 212.318 152

Unresponsive Between Groups .000 1 .000 .000 .999 Within Groups 217.325 151 1.439 Total 217.325 152


We were unable to reject Hypothesis 6: There is no significant difference in perceptions ofdisrespectful student behavior as measured by the MES between professors teaching at institutions

15


with a mandatory code-of-conduct and those without a mandatory code-of-conduct. ANOVAresults are shown in Table 8.

Table 8: Code of Conduct



Unresponsive Between Groups 5.600 2 2.800 2.004 .138 Within Groups 212.336 152 1.397 Total 217.936 154


We were unable to reject Hypothesis 7: There is no significant difference in perceptions ofdisrespectful student behavior as measured by the MES among professors with different ethnicity.ANOVA results are shown in Table 9.

Table 9: Ethnicity


Disagreeable Between Groups 6.768 4 1.692 1.235 .299 Within Groups 204.215 149 1.371 Total 210.984 153


Attendance Between Groups 4.752 4 1.188 .904 .463 Within Groups 195.714 149 1.314 Total 200.466 153

We rejected Hypothesis 8: There is a significant difference in perceptions of encounterswith disrespectful student behavior as measured by the MES between 103 administrators andbusiness faculty listed in the AACSB directory and the 54 business faculty teaching at HBCUinstitutions, with means of .05 for AACSB and -.06 for HBCU on factor two (Unresponsive). There

16


is a significant difference in perceptions of encounters with disrespectful student behavior asmeasured by the MES between 103 administrators and business faculty listed in the AACSBdirectory and the 54 business faculty teaching at HBCU institutions, with means of -.30 for AACSBand .54 for HBCU on factor three (Attendance). There appears to be opposite views on these twofactors. Administrators and faculty at AACSB schools seem to view an unresponsive studentencounter as more disrespectful than an HBCU faculty member. To the contrary, HBCU facultymembers view their encounters with students’ poor attendance to be significantly moredisrespectful than AACBS faculty and administrators do. MANCOVA results are shown in Tables10a and 10b.

Effect Value F Hypothesis df Error df Sig.Partial EtaSquared

Noncent.Parameter

ObservedPower(a)

Pillai's Trace .171 10.481(b) 3.000 152.000 .000 .171 31.444 .999

Wilks' Lambda .829 10.481(b) 3.000 152.000 .000 .171 31.444 .999

Hotelling's Trace .207 10.481(b) 3.000 152.000 .000 .171 31.444 .999

Roy's Largest Root .207 10.481(b) 3.000 152.000 .000 .171 31.444 .999

a) Computed using alpha = .05b) Exact statisticc) Design: Intercept+Status+OrgType

Table 10b: Tests of Between-Subjects Effects

Source Three Factors

Type IIISum ofSquares df

MeanSquare F Sig.

PartialEta

SquaredNoncent.Parameter

ObservedPower(a)

Model Disagreeable .175(b) 2 .088 .063 .939 .001 .127 .059

Unresponsive 12.943(c) 2 6.471 4.806 .009** .059 9.613 .791

Attendance 25.944(d) 2 12.972 11.395 .000*** .129 22.790 .992

a) Computed using alpha = .05; **Denotes significance at p< .01; *** denotes p< .001. b) R Squared = .001 (Adjusted R Squared = -.012)c) R Squared = .059 (Adjusted R Squared = .047)d) R Squared = .129 (Adjusted R Squared = .118)

DISCUSSION

We discuss our findings regarding faculty’s perceptions of their encounters withdisrespectful students’ behaviors in this section. The faculty reports that the most disrespectfulstudent behavior is students who come to class unprepared (not reading required materials, having

17


school supplies, and submitting required work on time); while students who miss several days ofclass without prior notice ranks second. These behaviors are indicative of students who wereunresponsive and who had problems with class attendance. These findings were not surprising, andwere reported by Charles (1999) and Seidman (2005).

The third and fourth most disrespectful behaviors were students who walked out of classwithout permission, and students who were not attentive during the lecture (sleeping, readingunrelated material, talking, browsing the Web, email of text messaging, etc.). (See Appendix A fora list of all reported behaviors and their mean rank.) In this technology-driven world, these findingsare not unexpected because students seem to be tied to their mobile phones and are constantlybrowsing the web, texting, etc. These students are too unresponsive and have problems with classattendance. However, the two least reported behaviors that faculty encountered were students whoopenly challenged their knowledge, and students who stormed out of class in anger.

Our analysis revealed that a full-time faculty member is more likely than a part-time facultymember to perceive an encounter with an unresponsive student as disrespectful. This is notunexpected because full-time faculty has more day-to-day interaction with students. They tend tobe involved with student organizations, and serve as role models and mentors for some students.Quddus, et al (2009) also reported a similar finding.

Our analysis also revealed that faculty and administrators from AACSB and HBCU schoolsdiffered in their perceptions of student disrespectful behavior as it relates to unresponsiveness andto attendance. AASCB members view an unresponsive student as more disrespectful than HBCUmembers. Perhaps some of the differences can be related to cultural differences and perceptions asnoted by Tom (1998). Alternatively, HBCU members view a student who has problems with classattendance as being more disrespectful than AACSB members. Perhaps the emphasis on a studentattending class at HBCU schools is attributable to the thought that students are exposed to the classmaterials through lecture, etc. rather than having to learn it on their own because they are not inclass.

Our study revealed that the faculty and administrators in our study find the students whoare most disrespectful are unresponsive and do not regularly attend classes. Students who aredisagreeable, i.e. they storm out of class in anger, argue with the professor or other class members,etc. were not found to be behaviors that faculty had regularly encountered.

RECOMMENDATIONS

To create a healthy learning environment in the classroom, we believe that faculty shouldbe in charge in the classroom. For this to happen, we encourage faculty to make their expectationsof classroom etiquette known to the students and discuss these from time to time for emphasis. Ifa faculty considers tardiness to be disruptive, the students should be made aware of this and points

18


taken off, if feasible. The incentives have to be appropriate. Poor behavior should be consistentlypenalized and good behavior rewarded. Therefore, we make the following recommendations.

Any college of business could adopt a set of “Expectations for Classroom Behavior” thatmany faculty members could share on the first day of the class. These explain the facultyexpectations of what is considered acceptable and unacceptable behavior, and provides theinstructor latitude to penalize poor behavior. The expectations should be discussed with thestudents, signed by the student, and filed.

When a student is unresponsive or has poor attendance, the instructor should privately meetwith the student to inform the student of the misconduct at the first sign, and then reinforce thedispleasure in a written reprimand. The faculty is also advised to seek the assistance of thedepartment head or the Dean’s office in case of extreme disruptive behavior. The professor mayalso contact the university’s assistant vice president in charge of disciplinary actions, or a similaradministrator if the student’s disrespectful behavior is extreme. Given these recommendations,there should be an improvement in the general student behavior in the classroom.

REFERENCES

A Loss of Respect (2004). Spectator, 294(9167), 7-7.

Boice, B. (1986). Faculty development via field programs for middle-aged, disillusioned faculty. Research in HigherEducation, 25, 115-135.

Boice, B. (1993). Primal origins and later correctives for midcareer disillusionment. In M. J. Finkelstein (Ed.),Developing senior faculty as teacher (New Directions in Teaching and Learning, 55, 33-41). San Francisco:Jossey-Bass.

Brown II, M. C. & J. E. Davis (2001). The Historically Black College as Social Contract, Social Capital, and SocialEqualizer. Peabody Journal of Education, 76(1), 31-49.

Brubacher, J. S. & Rudy, W. (1997). Higher education in transition: A history of American colleges and universities(Forth Edition.). U.S.A.: Transaction Publishers.

Buttner, H. E. (2004). How do we “dis” students?: A model of (dis)respectful business instructor behavior. Journal ofManagement Education, 28(3), 319-334, June.

Charles, C. (1999) Building classroom discipline. New York: Addison Wesley Longman, Inc.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika. 16, 297-334.

Cronbach, L. J. (1984). Essentials of psychological testing. New York: Harper and Row.

Devellis, R. (1991). Scale development. USA: Sage Publications.

19


Emerick, R. (1994). A conversation on classroom etiquette in introductory sociology courses. Teaching Sociology, 22,341-344.

Evers, W. J., Tomic, W., & Brouwers, A. (2004). Burnout among teachers: Students’ and teachers’ perceptionscompared. School Psychology International, 25(2), 131-148.

Guadagnoli, E. & Velicer, W. F. (1988). Relation of the sample size to the stability of component patterns.Psychological Bulletin, 103(2), 265-275.

Hatcher, L. (1994). A step-by-step approach to using the SAS(R) system for factor analysis and structural equationmodeling. Cary, NC: SAS Institute.

Harrison, E. C. (Spring 1972). Unrest on Black college campus. The Journal of Negro Education, 41(2), 113-120.

Hendrix, K. G. “She must be ‘trippin’”: The secret of disrespect from students of color toward faculty of color. NewDirections for Teaching and Learning, No. 110, published online at www.interscience.wiley.com.

Hessinger, R. (Autumn, 1999). The Most Powerful Instrument of College Discipline: Student Disorder and the Growthof Meritocracy in the Colleges of the Early Republic. History of Education Quarterly, 39(3), 237-262.

Jones, M. D. (Summer – Autumn, 1985). Student unrest at Talladega College. The Journal of Negro History, 70(3/4),73-81.

Kachigan, S. K. (1991). Multivariate statistical analysis. New York: Radius Press.

Kim, M. & Clifton, C. (June 2006). The Impact of Historically Black Colleges and Universities on the AcademicSuccess of African-American Students. Research in Higher Education, Vol. 47, No. 4, 399-427.

Meyers, S. (2003). Strategies to prevent and reduce conflict in college classrooms. College Teaching, 51(3), 94-98.

National Center for Education Statistics, Higher Education General Information Survey (August 2007). FallEnrollment in Institutions of Higher Education, 1980: and 1990 through 2006. Integrated PostsecondaryEducation Data System, downloaded from http://www.nces.ed.gov.

Nunnally, J. (1978). Psychometric theory. New York: McGraw-Hill.

Quddus, M. U., Bell, R. L., Bodie, N., Dyck, J., Rahman, S., Holloway, R., Desselle, B., & Till, A. (2009). Facultyperceptions and encounters with disrespectful student behavior. Academy of Educational Leadership Journal,13 (1), 1-18.

Seidman, A. (2005, spring). The learning killer: Disruptive student behavior in the classroom. Reading Improvement,42(1), 40-46.

Tom, G. (1998). Faculty and student perceptions of classroom etiquette. Journal of College Student Development,39(5), 515-516.

20


Williams, J.A. (1994). Classroom in conflict: Teaching controversial subjects in a diverse society. Albany, NY: SUNYPress.

Willie, C. V. and D. Cunnigen (1981). Black studies in higher education: a review of studies, 1965–1980. AnnualReview of Sociology, 7, 177-198.

APPENDIX A

Actual Scale Items on the MES Measuring Perceptions of Encounters with Disrespectful Students

At least two students have come to class unprepared (not reading required materials, etc.) 5.34 (1)

At least two students have missed several days of class without prior notice. 5.02 (2)

At least two students have walked out of the class without permission. 4.89 (3)

At least two students are not attentive during the lecture (sleeping, reading unrelated material, etc.) 4.52 (4)

At least two students talk to other students during my lectures. 4.23 (5)

At least two students have walked out of the class without permission. 3.69 (6)

At least two students choose not to participate in mandatory class assignments or activities 3.40 (7)

At least two students wear improper dress in my class (baseball hat, t-shirt with profanity, etc.). 3.31 (8)

At least two students have answered cell phones while class was in session. 2.69 (9)

At least two students in my class are argumentative with me and other students 2.60 (10)

At least two students have argued vehemently with me about an assignment or grade. 2.46 (11)

At least two students treat me and other students discourteously (interrupting, bad mouthing, etc.). 2.35 (12)

At least two students do not respond to my questions when I call on them. 2.24 (13)

At least two students address me by my first name in class. 2.05 (14)

At least two students have openly challenged my knowledge. 2.00 (15)

At least two students have stormed out of class in anger. 1.66 (16)

21


THE KEYS TO E-SERVICE RECOVERY:A FAST AND FAIR FIX

Daisy Wang, Minot State UniversityChia-Jung Lin, Robert Morris University

Kiattisak (Pong) Phongkusolchit, University of Tennessee at Martin

ABSTRACT

The growth of business-to-customer (B2C) electronic commerce unraveled tremendousopportunities for service providers. Along with this opportunity comes the challenge of identifyingand rectifying service failures. Despite this growing interest on electronic service research, theempirical evidence in the literature is still evolving. The purpose of this study is to empiricallyexamine the effect of the speed and magnitude of online service recovery activities on customersatisfaction, loyalty and positive Word-of-Mouth (WOM) in a B2C context. The results of our studystrongly support a positive impact of the speed and magnitude of online service recovery on post-recovery customer behaviors. Research and practical implications are discussed.

Keywords: online service recovery, service failures, electronic commerce, speed of recovery,magnitude of recovery

INTRODUCTION

The growth of business-to-customer (B2C) electronic commerce unraveled tremendousopportunities for service providers. The convenient access of Internet technology by millions ofconsumers provides an excellent ground for the emergence of electronic retailers such asamazon.com and e-bay.com and online operation of brick-and-mortar stores, such as Walmart.comand Sears.com. Regardless of the type of service they provide, service providers sooner or laterexperience some degree of service failure. Research suggests that it costs up to five times as muchto attract new consumers than it does to keep current consumers satisfied (Hart, Heskett, & Sasser,1990; Maxham, 2001). As such, identifying and rectifying these service failures is crucial forachieving both consumer retention and overall competitiveness.

Service recovery studies have identified a number of important factors that are associatedwith the recovery process. Specifically, the antecedents of service recovery activities such as thedegree of consumer’s loyalty, perceived service quality and severity of failure are attributed to have

22


an impact on the success of recovery activities and consumer satisfaction (Bitner, 1990; Kelley &Davis, 1994; Miller, Craighead, & Karwan, 2000). The extant literature also provides evidence thatsuccessful service recovery activities involve both psychological and tangible compensations (Bell& Ridge, 1992; Clark, Kaminiski, & Rink, 1992), such as promptness, fairness, and monetarycompensation.

However, most of the findings in the literature are based on service recovery efforts in the“traditional” service setting (i.e., service provided with direct physical contact between theconsumer and service provider). Consequently, there is a limited understanding of online servicerecovery activities in a B2C setting (Holloway & Beatty, 2003). Examining the impact of B2Cservice recovery activities is particularly important given the phenomenal growth of electroniccommerce both in the form of business-to-business (B2B) and business-to-customer (B2C)transactions. Online service recovery is significantly different from brick-and-mortar settingprimarily due to customers’ perceived insecurity of internet transactions and the lack ofinterpersonal interaction (Forbes, Kelley, & Hoffman, 2005b; Holloway, Wang, & Parish, 2005).

We believe that due to the evolving nature of online service recovery literature, there is alack of an empirical study on the nature of web-based service recovery activities in relation tocustomer satisfaction, loyalty, and positive word-of-mouth (Holloway & Beatty, 2003). This studyis intended to fill the gap of the literature and specifically to empirically examine the effect of thespeed and magnitude of online service recovery activities on customer satisfaction, loyalty andpositive word-of-mouth (WOM) in a B2C context.

THEORETICAL BACKGROUND AND HYPOTHESIS DEVELOPMENT

Web-Based Service Recovery

A service failure can be defined as “… any service mishap or problem that occurs duringa consumer’s experience with a firm” (Maxham, 2001, p. 11). It occurs when a consumer’sperception of actual service experience falls below the expected level (Holloway & Beatty, 2003).In an exploratory study of online retailers, Holloway and Beatty (2003) identified the most commontypes of online service failure including delivery, website design, payment and security problems.Service recovery in general refers to “… those actions designed to resolve problems, alter negativeattitudes of dissatisfied consumers and to ultimately retain these consumers” (Miller, et al., 2000,p. 388).

Since service failures are inevitable in both traditional and web-based service providers,effective service recovery efforts are essential for maintaining consumer satisfaction and loyalty.These efforts can be especially critical for web-based service providers (particularly those firmsproviding B2C services) when facing the challenge of resolving service failure situations withouta face-to-face interaction with the consumer. This is a significant impediment given that providing

23


both psychological (e.g., empathy and apology) and tangible compensation as part of a full-fledgedservice recovery effort (Boshoff, 1999; Miller, et al., 2000). Moreover, consumers’ level of trustin online transactions as well as complaining behavior in the event of a service failure to somedegree influence the web-based service recovery process (Hoffman, Novak, & Peralta, 1999; Lee& Turban, 2001). Most importantly, the relatively low switching cost for consumers demonstratessignificant pressure on online service providers in terms of resolving service problems promptlyand effectively in order to maintain consumer loyalty. As such, this study empirically examinesthe impact of the speed and extent of service recovery effort on consumer satisfaction, consumerloyalty and positive word-of-mouth (WOM) in the B2C environment.

The extant literature further provides evidence for the declining post-service consumersatisfaction as a result of low levels of service recovery efforts (Goodwin & Ross, 1992; Kelley &Davis, 1994; McCollough & Berry, 1996). Researchers have adopted justice theory extensivelyto explain the process of service recovery activities and its consequences on consumers (Wirtz &Mattila, 2004). Distributive justice refers to the consumer’s perception of what he or she receivesas the outcome of the service recovery process(McColl-Kennedy & Sparks, 2003). A free air ticketfor a delayed or cancelled flight from an airline company, or a free drink voucher for a slow mealservice in a restaurant belongs to this category. Procedural justice in the service climate, on theother hand, is defined as the actual process that was used by the service provider to resolve theproblem. Hence, it emphasizes the process more than the outcome (Smith, et al., 1999).Interactional justice deals with the way the service failure situation was handled between theconsumer and service provider. Whether or not the service provider treats the consumerexperiencing the service failure with sensitivity, empathy, compassion, dignity, and respect isconsidered as examples of interactional justice (Sparks & Mccoll-Kennedy, 2001).

Speed of Web-Based Service Recovery

One of the important efforts of service recovery is the degree of promptness. Especially inelectronic service delivery settings, speed of recovery could play a crucial role in maintainingconsumer satisfaction and loyalty in condition of relatively low switching costs for e-shoppers(Schweikhart, Strasser, & Kennedy, 1993; Spreng, Harrell, & Mackoy, 1995). In the light ofprocedural and interactional justice, a higher level of perceived promptness of online servicerecovery might help e-business to retain the customers with a result of improved customersatisfaction, loyalty, and even positive word-of-mouth. Generally speaking, most online shoppersare reluctant to continue doing business with an online retailer where they experience a servicefailure unless there is a prompt service recovery effort from the service provider. For instance,Hart, et al. (1990) noted that a service failure tends to be resolved successfully if the problem issolved promptly. Similarly, Wirtz and Mattila (2004) also found that the speed of service-providers

24


responding to a customer’s complaint could serve as an indicator of organizational efficiency thatspeed of recovery is positively related to the overall perceived quality of the service providers.

Some other studies have also discussed consumer satisfaction the equity theory astheoretical framework. Customers would perceive differently based on equity of the situations andoutcomes or compensation from the service providers. For example, Lapidus and Pinkerton (1995)adopt equity theory to examine various customer complaint situations. Their study found thatsubjects would perceive more resentment in an inequitable and low outcome situation, wouldperceive fairer and favorable to service providers offering a higher compensation, and wouldperceive more guilt in inequitable, high outcome situations. In sum, based on the equity theoryperspective, consumer satisfaction is usually positively related to consumers’ perceived fairness(Clemmer & Schneider, 1996; Oliver & Swan, 1989). In other words, perceived consumersatisfaction will increase as firms “fairly” and successfully resolve their service failures.Accordingly, perceived consumer satisfaction following a service failure is linked to perceivedfairness in the recovery process (Maxham, 2001). In addition, a slow response to web-basedservice failures such as incorrect order item, incomplete billing, poor consumer service could resultin lower consumer loyalty (Hart, et al., 1990). Hence, it is hypothesized that:

Hypothesis 1a: The speed of web-based service recovery is positively related topost-recovery customer satisfaction.

Hypothesis 1b: The speed of web-based service recovery is positively relatedto post-recovery customer loyalty (repurchase intentions).

Hypothesis 1c: The speed of web-based service recovery is positivelyrelated to post-recovery customers’ positive word-of-mouth(WOM).

Magnitude of Service Recovery

Effective service recovery efforts involve not only prompt communication but also adequateand tangible compensation, which is also known as the magnitude of service recovery (Miller, etal., 2000). Prior studies on service recovery activities indicate that, overall, two major elementscritical to ensure the effectiveness and completeness of service recoveries are psychological andtangible efforts (Miller, et al., 2000; Schweikhart, et al., 1993). The psychological elements mayinclude empathizing and apologizing (Bell & Ridge, 1992; Zemke, 1994). Tangible recovery effortis to compensate for real and perceived damage (Zemke, 1994) that are aimed at providing a fairrestitution for consumers’ losses and inconveniences of the service failure. A secondary intent may

25


be to provide value-added atonement (see, e.g. Clark, et al., 1992; Hoffman, et al., 1995, Miller etal., 2000).

The magnitude of service recovery (i.e., the extent of tangibles provided) could result inhigher consumer satisfaction since the consumer perceives that the service provider went “extramile” to provide fair restitution. For instance, Clark et al. (1992) found that 91% of subjects in theirstudy who received a “little extra” after a service failure intended to stay loyal to the specificservice provider. This is consistent with the basic notion of expectation disconfirmation theory thatdefines satisfaction as exceeding customers’ expectation (Andreassen, 2000; Oliver & Swan, 1989).In addition, the magnitude of service recovery in electronic transactions could also positively affectthe spread of positive word-of-mouth (WOM) by the consumer. Prior studies have found a positivelink between service recovery effort and WOM (Blodgett, Granbois, & Walters, 1993; Bloodget,Hill, & Tax, 1997). Hence, it is hypothesized that:

Hypothesis 2a: The magnitude of web-based service recovery positivelyinfluences post-recovery customer satisfaction.

Hypothesis 2b: The magnitude of web-based service recovery is positivelyrelated to post-recovery customer loyalty (repurchaseintentions).

Hypothesis 2c: The magnitude of web-based service recovery has a positiverelationship with post-recovery customers’ positive word-of-mouth.

METHODOLOGY

Research Setting

A scenario-based experiment was used to gather data from respondents. A scenario-drivenservice research not only enables the researcher to gather important service recovery informationbut also manipulates key study variables for empirical testing (Harris, Grewal, Mohr, & Bernhardt,2006; Weiner, 2000). We believe the use of scenarios in service research is appropriate at least forthe following two reasons. First, carefully manipulated scenarios help researchers capture the widerange of customer responses to various service recovery activities (Smith, et al., 1999). Second,the use of scenarios reduces the potential for respondents’ memory-bias, which is often a commondrawback found in self-reported studies (Smith, et al., 1999; Smith, Bolton, & Wagner, 1998; Wirtz& Mattila, 2004).

26


With scenarios, service recovery manipulations could be better controlled, and potentialethical problems related to the actual experience of service failures could be avoided. The scenarioused in this study was constructed to manipulate the speed and magnitude of online servicerecovery activities. We used a hypothetical scenario with two levels (low/high) of our independentvariables (i.e., speed and magnitude of service recovery) in an online book retailer setting.Although the independent variables could be treated as a continuous variable, we chose to presentthem as categorical variables in order to distinctly capture the variance in our manipulations.

Sample

In order to test the above hypotheses, data were gathered using a scenario-based experimentwith 151 undergraduate business students at a large Midwestern pubic university. We believecollege students are suitable for this study based on the fact that their lifestyle involves using theInternet a lot not only for entertainment but also for shopping purposes. In addition, they areexpected to have more confidence and aptitude in using technology in general and online shoppingin particular. 63% of the respondents are male and mainly between the ages of 20-29 (80%). 65%of the respondents are Caucasians. In terms of online shopping behavior, fifty-one percent of therespondents indicated they shop online every month and forty percent of them admitted they spentbetween $50 and $250 per online purchase.

Measures

This research follows Maxham’s (2000) study to use customer satisfaction, purchase intentand word-of-mouth as measures of post-recovery firm performance.

Post-recovery satisfaction

Post-recovery consumer satisfaction was measured using four items adopted from Boshoff’s(1999) instrument. Although original instrument includes six items (communication, empowerment,feedback, atonement, explanation and tangibles) measuring post-recovery consumer satisfaction,we only select four items as measures of post-recovery satisfaction due to the nature of onlinesetting. The Cronbach’s Alpha for the four items was 0.96 indicating a high reliability. The fouritems were averaged to form a composite measure of post-recovery satisfaction.

Customer loyalty

Three items were used to measure customer loyalty (i.e., repurchase intentions) anchoredon a five point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The items used

27


in our study were adapted from those of Maxham’s (2000) study. Since most of the studies in theliterature focused on the brick-and-mortar setting, the items were modified to fit the scope of ourstudy focusing on online setting. The Cronbach’s Alpha for the three items was 0.92 showing ahigh reliability. The three items were averaged to form a composite measure of customer loyalty.

Positive word-of-mouth (WOM)

We measured positive WOM using two items anchored on a five point Likert scale rangingfrom 1 (strongly disagree) to 5 (strongly agree). These items are: “I am more likely to say positivethings about the online book retailer to other people” and “I am more likely to encourage friendsand others to do business with this online service provider” (Maxham, 2000). The Cronbach’sAlpha for these two items was 0.95. Similar to the other dependent variables, we averaged the twoitems to create a composite measure of positive WOM.

DATA ANALYSIS AND RESULTS

Preliminary Study

Before conducting the main study, we pre-tested the measures of post-recovery customersatisfaction, loyalty and positive WOM using a separate group of 189 undergraduate students fromthe same institution. We presented a hypothetical online book retailer scenario followed bymeasurement questions regarding post-recovery customer satisfaction, loyalty, positive WOM, anddemographic information.

No online retailer name was mentioned in the scenario to eliminate potential respondentbias. The respondents were asked to imagine a situation that they bought a textbook from an onlinebook retailer right before the semester began. They did not receive the book on the date of deliveryas promised in the confirmation email sent by the online book retailer. Therefore, they e-mailedthe online book retailer to complain the unfulfilling book transaction. For pre-testing purposes, wegave the respondents a scenario with a single manipulated service recovery (low vs. high speed ofrecovery) by a hypothetical online book retailer.

Manipulation Check and Reliability Test

We then performed a manipulation check on the low vs. high service recovery. The resultsindicate that our scenario is perceived as planned by respondents and that the manipulation issignificantly different between the two groups (F (1, 187) = 5.75, p < 0.05). We also asked therespondents to what extent they think the scenarios are realistic and believable on a five-point scale

28


ranging from 1 to 5. Accordingly, the results indicate that most of the respondents think thescenario is indeed believable (Mean = 3.94) and realistic (Mean = 3.86).

Main Study

The participants were asked to read the unfulfilling book transaction description similar tothe one in preliminary study, followed by one of the four manipulated service recovery situations(i.e., low speed-low magnitude, high speed-high magnitude, high speed-low magnitude and lowspeed-high magnitude of recovery). This manipulation result could be analyzed by a 2X2Multivariate Analysis of Variance (MANOVA) with randomized equal cell assignments.MANOVA is suitable for hypothesis-testing in this study since we are interested in testing theeffect of two categorical independent variables on three criterion variables (Hair, Black, Babin,Anderson, & Tatham, 2006). Prior research has shown that customers’ online purchase experiencecan affect their satisfaction with service recovery activities (Holloway, et al., 2005; Wirtz &Mattila, 2004).

Results

The positive relationships between the speed of web-based service recovery and post-recovery customer behaviors predicted in the first set of hypotheses are supported (Wilks λ = 0.59,F = 138.67, P < 0.01). More specifically, Hypothesis 1a (H1a) proposed a positive relationshipbetween the speed of web-based service recovery and post-recovery customer satisfaction. Theresults provide support for H1a (F (1, 598) = 405.953, P < 0.001). Hypothesis 1b (H1b) stated apositive relationship between the speed of online service recovery and customer loyalty (repurchaseintention) while hypothesis 1c (H1c) predicted a positive relationship between the speed of onlineservice recovery and post-recovery positive WOM. The results of our analysis provided supportfor both H1b (F (1, 598) = 168.293, p< 0.001) and H1c (F (1,598) = 138.976, p< 0.001). Table 1summarizes the MANOVA results on service recovery performance.

Table 1: Statistical Summary of MANOVA on Customer Satisfaction, Loyalty, and WOM

Analysis Multivariate Univariate Univariate Univariate

IVs \ DVs MANCOVA Customer Satisfaction Customer Loyalty WOM

Speed of Service Recovery 138.67*** 405.95**** 168.293**** 138.976****

Magnitude of Service Recovery 178.85*** 529.11**** 197.998**** 216.192****

Speed X Magnitude 32.18**

* Significant at 0.1 ** significant at 0.05 ***significant at 0.01 **** significant at 0.001

29


Figure 2 shows the overall effect of the customers’ perceived speed of service recovery onsatisfaction, loyalty, and positive WOM. While the results on all three service recovery performancemeasures were significant, more dispersion of post-recovery customer satisfaction is shown betweenslow and fast service recovery manipulation.

Hypotheses 2a-2c proposed a positive relationship between the magnitude of servicerecovery, post-recovery satisfaction, loyalty and WOM respectively. The results show a strongsupport for the relationship between the magnitude of web-based service recovery and overall post-recovery customer behaviors (Wilks λ = 0.53, F = 178.85, P < 0.01). The univariate results providesupport for the positive relationship between the magnitude of web-based service recovery and post-recovery customer satisfaction (F (1, 598) = 529.105, p < 0.001), loyalty (F (1, 598) = 197.998, P < 0.001)as well as positive WOM (F (1, 598) = 216.192, P < 0.001).

Figure 2: The Effect of Speed of recovery on Satisfaction, Loyalty and WOMa

Slow vs. Fast Speed of Service Recovery

0

2

4

6

8

10

12

Post-Recovery Satisfaction Loyalty WOM

Mean Scores

SlowFast

a Mean Scores for variables are based on summed-item scores

Figure 3 presents the mean scores of post-recovery customer satisfaction, loyalty and positiveWOM across low and high magnitudes of service recovery. Even though it was not part of ouroriginal hypothesized relationships, we also observed a significant interaction between the speed andmagnitude of online service recovery activities (Wilks λ = 0.86, F = 32.18, P < 0.01). This findingindicates that consumers would perceive these two types of recovery activities interdependently orthat firms that respond fast to the service failure usually respond with magnitudes.

30


Figure 3: The Effect of Magnitude of recovery on Satisfaction, Loyalty and WOMa

Low vs. High Magnitude of Service Recovery

0

2

4

6

8

10

12

14

Post-Recovery Satisfaction Loyalty WOM

Mean Scores

LowHigh

a Mean Scores for variables are based on summed-item scores

DISCUSSION AND CONCLUSION

Implications

The purpose of this study is to investigate whether the speed and magnitude of web-basedservice recovery significantly affect post-recovery customer satisfaction, loyalty and positive WOM.As presented above, the results provide a strong support for the aforementioned hypotheses. Ingeneral, the results of this study indicate that online shoppers experiencing a service failure situationare more likely to be satisfied with the online service provider’s solution if it is done in a moretimely manner and exceeding their expected compensation. The results also show that onlineshoppers are more likely to be loyal and spread positive WOM if they perceive that they receiveda speedy and fair resolution to their service problems.

These results are consistent with the findings in the extant literature (Holloway, et al., 2005;Wirtz & Mattila, 2004). For instance, Wirtz and Mattila (2004), in their study of 187 subjects usinga restaurant scenario, found that the speed of recovery was positively related to higher repurchaseintention and lower negative WOM. They also found a significant three way interaction amongspeed, apology and compensation on post-recovery customer satisfaction. However, their overallresult suggests a stronger combination effect of immediate service recovery with apology increasepost-recovery satisfaction. We believe that the results of our study contribute to the ongoingresearch particularly in the area of online service recovery by highlighting the role of offering aprompt service recovery as it might be a clue regarding service efficiency (McColl-Kennedy &Sparks, 2003; Smith, et al., 1999). More specifically, the role of procedural (i.e. speed of recovery)

31


and distributive (i.e. magnitude of recovery) justice seem to suggest particular relevance to onlineservice recovery activities.

Our findings also offer important insights on the roles of speed and magnitude of web-basedservice recovery activities for online businesses. With abundance of service options, low entrybarriers and minimal switching costs (Holloway & Beatty, 2003), today’s web-based businesses ingeneral and online retailers in particular have to focus their resources on building customerrelationship in a fashion of speedy and fair resolution. As such, enhancing the speed and magnitudeof web-based service recovery not only benefits B2C firms to maintain customer loyalty but alsoenables them to use existing customers as efficient marketing mediums via positive WOM.

Suggestions for Future Research

The findings of this study shed lights on a number of important avenues for future research.For instance, although we examined the role of the speed and magnitude of web-based servicerecovery as indicators of customers’ perceived procedural and distributive justice, more research isneeded on the implications of interactional justice in online service recovery process. Given thephenomenal growth of instant messaging and chatting, future research in this area should examinehow online service providers use such features for effective and prompt service recovery. Anotherfuture research possibility could be examining the online service recovery activities difference, ifany, between companies with a pure online business model and those with a combination of onlineand brick-and-mortar establishments. Future research could also investigate the effect of consumers’trust and firm’s brand awareness or reputation on online service recovery effectiveness.

Conclusion

There is an increasing scholarly interest in both online service quality and service recoveryresearch (Forbes, Kelley, & Hoffman, 2005a; Holloway, et al., 2005; Parasuraman, Zeithmal, &Malhotra, 2005). Despite the growing importance of online service research, the empirical evidencein the literature is still evolving (Holloway & Beatty, 2003). This study contributes to a betterunderstanding of web-based service recovery effort and its repercussions on customer satisfaction,loyalty and positive WOM.

REFERENCES

Andreassen, T. W. (2000). Antecedents to Satisfaction with Service Recovery. European Journal of Marketing, 34(1/2),156-175.

Bell, C., & Ridge, K. (1992). Service Recovery for Trainers. Training and Development, 46(5), 58-63.

32


Bitner, M. J. (1990). Evaluating the service encounters: the effects of physical Surroundings and employee responses.Journal of Marketing, 54(1), 95-106.

Blodgett, J. G., Granbois, D. H., & Walters, R. G. (1993). The effects of perceived justice on Complainant's NegativeWord-of-mouth Behavior and Repartronage Intentions. Journal of Retailing, 69(4), 399-427.

Bloodget, J. G., Hill, D. J., & Tax, S. S. (1997). The Effects of Distributive, Procedural, and Interactional Justice on postcomplainant behavior. Journal of Retailing, 73(2), 185-210.

Boshoff, C. (1999). RECOVSAT: an instrument to measure satisfaction with transaction-specific service recovery.Journal of Service Research, 1(3), 236-249.

Clark, G. L., Kaminiski, P. F., & Rink, D. R. (1992). Consumer complaints: advice on how Companies should respondbased on an empirical study. The Journal of Services Marketing, 6(1), 41-50.

Clemmer, E. C., & Schneider, B. (1996). Fair Service. In T. A. Swartz, D. E. Bowen & S. W. Brown (Eds.), Advancesin Services Marketing and Management (Vol. 5, pp. 109-126). Greenwich, CT: JAI Press.

Forbes, L. P., Kelley, S. W., & Hoffman, K. (2005a). Typologies of E-Commerce Retail Failures and RecoveryStrategies. Journal of Services Marketing, 19(5).

Forbes, L. P., Kelley, S. W., & Hoffman, K. D. (2005b). Typologies of E-Commerce Retail Failures and RecoveryStrategies. Journal of Services Marketing, 19(5), 280-292.

Goodwin, C., & Ross, I. (1992). Consumer responses to service failures: an influence of Procedural and Interactionalfairness perceptions. Journal of Business Research, 25(2), 149-163.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate Data Analysis (6th ed.).Upper Saddle River, NJ: Pearson Education, Inc.

Harris, K. E., Grewal, D., Mohr, L. A., & Bernhardt, K. L. (2006). Consumer Responses to Service Recovery Strategies:The Moderating Role of Online versus Offline Environment. Journal of Business Research, 59, 425-431.

Hart, C. W., Heskett, J., & Sasser, W. (1990). The profitable Art of Service Recovery. Harvard Business Review,August, 148-156.

Hoffman, D. L., Novak, T. P., & Peralta, M. (1999). Building consumer trust online. Communications of the ACM, 42(4),80-85.

Holloway, B. B., & Beatty, S. (2003). Service Failure in Online Retailing: A Recovery Opportunity. Journal of ServiceResearch, 6(1), 91-105.

Holloway, B. B., Wang, S., & Parish, J. T. (2005). The Role of Cumulative Online Purchasing Experience in ServiceRecovery Management. Journal of Interactive Marketing, 19(3), 55-66.

33


Kelley, S. W., & Davis, M. A. (1994). Antecedents to consumer expectations for service Recovery. Journal of Academyof Marketing Sciences, 22(1), 52-61.

Lapidus, R. S., & Pinkerton, L. (1995). Customer complaint situations: An equity theory perspective. Psychology andMarketing, 12(2), 105-122.

Lee, M. K. O., & Turban, E. (2001). A Trust Model for Consumer Internet Shopping. International Journal of ElectronicCommerce 6(1), 75-91.

Maxham, J. G. (2001). Service recovery's influence on consumer satisfaction, positive Word-of-mouth, and purchaseintentions. Journal of Business Research, 54(1), 11-24.

McColl-Kennedy, J. R., & Sparks, B. A. (2003). Application of Fairness Theory to Service Failures and ServiceRecovery. Journal of Service Research, 5(3), 251-266.

McCollough, M. A., & Berry, L. L. (1996). A conceptual model and empirical investigation of consumer satisfactionafter service failure and recovery. In R. T. Rust & R. L. Oliver (Eds.), Frontiers in Services. A Conference co-sponsored by the American Marketing Association and the Center for Services Marketing and VanderbiltUniversity. Chicago: American Marketing Association.

Miller, J. L., Craighead, C. W., & Karwan, K. R. (2000). Service recovery: a framework and Empirical investigation.Journal of Operations Management, 18(4), 387-400.

Oliver, R., & Swan, J. E. (1989). Consumer perceptions of interpersonal equity and satisfaction In transactions: a fieldsurvey approach. Journal of Marketing, 53(2), 21-35.

Parasuraman, A., Zeithmal, V., & Malhotra, A. (2005). E-S-QUAL: A multiple-item scale for Assessing electronicservice quality. 7(3), 213-233.

Schweikhart, S. B., Strasser, S., & Kennedy, M. R. (1993). Service recovery in health Services organizations. Hospitaland Health Administration, 38(1), 3-21.

Smith, A. K., Bolton, R. N., & Wagner, J. (1999). A Model of Customer Satisfaction with Service Encounters InvolvingFailure and Recovery. Journal of Marketing Research, 36(3), 356-372.

Smith, A. K., Bolton, R. N., & Wagner, J. A. (1998). A model of consumer satisfaction with Service encountersinvolving failure and recovery. Marketing Science Institute.

Sparks, B. A., & Mccoll-Kennedy, J. R. (2001). Justice strategy options for increased customer satisfaction in a servicesrecovery setting. Journal of Business Research, 54(3), 209-218.

Spreng, R. A., Harrell, G. D., & Mackoy, R. D. (1995). Service recovery: impact on Satisfaction and intentions. Journalof Services Marketing, 9(1), 15-23.

Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences:implications for relationship marketing. The Journal of Marketing, 62(2), 60-76.

34


Weiner, B. (2000). Intrapersonal and Interpersonal Theories of Motivation from an Attributional Perspective EducationalPsychology Review, 12(1), 1-14.

Wirtz, J., & Mattila, A. (2004). Consumer Responses to Compensation, Speed of Recovery and Apology after a ServiceFailure. International Journal of Service Industry Management, 15(2), 150-166.

Zemke, R. (1994). Service recovery. Executive Excellence, 11(9), 17-18.

35


ARE S&P 500 INDEX MUTUAL FUNDSCREATED EQUALLY?

John Cresson, Southeastern Louisiana University

ABSTRACT

Standard and Poor’s (S&P) 500 Index mutual funds are designed to replicate the S&P 500Index, and therefore, should move very closely with the S&P 500 Index. They are supposed to behomogeneous goods. In this study, I analyze the performance of S&P 500 Index mutual funds. Thegreat appeal of researching Index funds is that there are limited opportunities in the equity marketsto study homogeneous investment vehicles. This study differs from previous papers in a number ofways. Unlike other researchers who use Morningstar exclusively as their data source, I employeetwo popular mutual fund data sources to conduct my analysis, Morningstar and Lipper. Also, Ianalyze both tracking and performance, and examine possible influences on performance. I presentevidence that tracking measures are very high and that managers do a good job of tracking the S&P500 Index. However, despite the high tracking measures, S&P 500 Index funds clearlyunderperform the S&P 500 Index. I present evidence that funds with lower expenses and fundswhich are more established have higher returns. Also, there is evidence that returns are positivelycorrelated with past returns.

INTRODUCTION

Standard and Poor’s (S&P) 500 Index mutual funds have become enormously popular withinvestors. Chen, Noronha, and Singal (2005) argue that part of the popularity results from theevidence that actively managed mutual funds do not consistently outperform index mutual funds.S&P 500 Index mutual funds are designed to replicate the S&P 500 Index. Frino, Gallagher,Neubert, and Oetomo (2004) note that the management of index portfolios is theoreticallystraightforward. Even though the management of index portfolios may be straight forward in theory,Frino and Gallagher (2001), and Elton, Gruber, and Busse (2004) note that the actual practice is farmore complicated.

Despite the popularity of S&P 500 Index mutual funds, Frino and Gallagher note thatempirical research addressing index funds is scarce. In this study, I analyze the performance ofS&P 500 Index mutual funds. The great appeal of researching Index funds is that there are limitedopportunities in the equity markets to study homogeneous investment vehicles. This study differsfrom previous papers in a number of ways. Unlike other researchers who use Morningstar

36


exclusively as their data source, I employee two popular mutual fund data sources to conduct myanalysis, Morningstar and Lipper. Also, I analyze both the tracking and performance of S&P 500Index mutual funds, and the possible influences on performance. (For studies that analyze tracking,see Frino & Gallagher (2001), Elton, Gruber & Busse (2004), and Frino, Gallagher & Neubert-Oetomo (2004).)

I present evidence that tracking measures are very high and that managers do a good job oftracking the S&P 500 Index. However, despite the high tracking measures, S&P 500 Index fundsclearly underperform the S&P 500 Index. Elton, Gruber, and Busse note that expenses may accountfor a portion of the differential performance between index funds. Presenting evidence consistentwith their argument, I find that funds with lower expenses have higher returns. I also find a positive,significant relationship between the return and the age of a fund. This is evidence that older, moreestablished funds have higher returns. Elton, Gruber, and Busse also note that investors shouldconsider if they can predict the returns that a fund will generate in the future. An investor who isconcerned about future returns will likely study past returns when picking funds. I present evidencethat returns are positively correlated with past returns.

Since S&P 500 Index funds are supposed to be homogeneous goods, investors may assumethat they are all the same. Even index fund managers admit that high expenses exist because thefunds feel they can get away with it. “Investors usually don’t think to compare our expenses. Theythink we’re all the same” according to a manager of a high-expense fund (“Up Front,” BusinessWeek, April 14, 1997, McGraw-Hill Companies). This paper presents evidence that there arevariations in returns across S&P 500 Index mutual funds and that higher expenses reduce returns.With so many index funds to analyze, investors may simply be overwhelmed with information. Thispaper provides research to simplify the analysis for investors.

S&P 500 INDEX MUTUAL FUNDS

S&P 500 Index mutual funds are designed to replicate the performance of the S&P 500Index, a market-value-weighted index of five hundred stocks that are traded on the New York StockExchange, the American Stock Exchange, and the NASDAQ National Market (S&P 500 1999/2000Directory). Since S&P 500 Index mutual funds are designed to replicate the S&P 500 Index, theyshould perform like the S&P 500 Index.

Even though the management of index portfolios may be straight forward in theory, actualpractice is far more complicated. Elton, Gruber, and Busse argue that even if index mutual fundsdid not charge expenses, their returns would probably differ from the Index’s returns. For example,it costs money to run an index mutual fund, while a pure index is a fictitious paper portfolio thatincurs no expenses. Also, the composition of the Index frequently changes. In 2001 alone, therewere 30 additions and 30 deletions to the S&P 500 Index, with at least one change occurring inevery month of the year. In 2002 there were 24 additions and 24 deletions occurring in months other

37


than March, April and October. In 2003 there were 9 additions and 9 deletions occurring inFebruary, March, April, July, August, September, and November (Source: Standard and Poor’s).

The goal of an S&P 500 Index mutual fund manager is to replicate the performance of theS&P 500 Index given market constraints. This is demonstrated by information provided byMorningstar and the mutual funds themselves for a sample of “S&P 500” mutual funds. Forexample, one fund seeks to provide investment results that correspond to the total return of the S&P500, while another fund seeks growth of capital and income that correspond to the investment returnof the S&P 500 index. The second fund will primarily invest in common stocks that comprise theindex.

DATA

In this paper, I employee two popular mutual fund data sources to conduct my analysis,Morningstar and Lipper. Construction of the Morningstar data set starts with a list of mutual fundsmonitored by Morningstar in 2003 that are entitled as “S&P 500” mutual funds. This leads toeighty-one funds. From this list, funds are eliminated if I cannot verify that they are indeed S&P500 Index funds or if they are enhanced or hedge S&P 500 Index funds. The final Morningstarsample consists of forty-two S&P 500 Index mutual funds.

The Lipper dataset consists of mutual funds monitored by Lipper that they classify as “S&P500 Index” mutual funds effective February 2004. This results in 172 funds. Lipper provides thethree-year, five-year and ten-year returns of the funds that they monitor. This leads to 151, 100, and39 S&P 500 Index mutual funds that are available for analyses that have returns for three, five, andten years, respectively. Lipper also provides a ranking of S&P 500 Index mutual funds in terms ofmeeting certain goals, such as expenses. The scores presented by Lipper are derived from complexformulas that scrutinize funds based on clear criteria with each fund being ranked against its peers.

The descriptive statistics of the S&P 500 Index mutual funds as collected from Morningstarare presented in Table 1. The average fund has more than 97 percent of its holdings in stock, andis relatively young (an average inception date of 1997). The average S&P 500 Index mutual fundhas more than one billion dollars in total assets, with one fund having as little as 13 million dollarsin assets and another fund having more than 6.8 billion dollars in assets. The average expense ratioof an S&P 500 Index fund is .5054 percent, with a low of .13 percent and a high of 1.5 percent.

38


Table 1: Descriptive Statistics of S&P 500 Index Mutual Funds

Mean Minimum Maximum

Total Assets ($Mill) 1,004 13 6,897

Expense Ratio (%) 5054 .13 1.5

Inception Date 1997 1985 2002

Percentage in Stock 97 93 100

This table presents the mean, minimum, and maximum of the total assets, expense ratio, inception date, andpercentage of the fund’s assets invested in stocks of S&P 500 Index mutual funds employing Morningstar data.

TRACKING RESULTS OF S&P 500 INDEX MUTUAL FUNDS

According to Elton, Gruber, and Busse, investors should care about how closely a fund tracksthe index. Although several measures of tracking error are presented in the literature, Chen,Noronha, and Singal (2005) note that there is no universally accepted definition. Elton, Gruber, andBusse employ the coefficient of determination, R2 of an ordinary least squares regression, to measuretracking. R2 of the regression model indicates the closeness to which the index fund mimics the S&P500 Index. According to Morningstar, R2 “reflects the percentage of a fund’s movements that areexplained by movements in its benchmark index. An R2 of 100 means that all movements of a fundare completely explained by the movements in the index. Thus, index funds that invest only in S&P500 stock will have an R2 very close to 100.” R2 is usually expressed as ranging from 0 to 1.However, Morningstar expresses R2 as having a range from 0 to 100 (in percentages). Morningstarprovides the R2 values of the mutual funds they monitor. R2 of each S&P 500 Index Mutual Fundis displayed in Table 2. According to Morningstar, every fund in my sample (with an available R2)has an R2 of 100. (Since Morningstar calculates R2 over a three-year period, R2 is available forthirty-five of the funds in my sample.) The mean R2 of the S&P 500 Index mutual funds is 100, witha standard deviation of 0. Thus, employing Morningstar data, every fund perfectly tracks the Index.My findings are consistent with the results of other studies that employ monthly returns indetermining R2 values of S&P 500 Index mutual funds. Considering how difficult it is to trulyreplicate the S&P 500 Index, these high values of R2 are very encouraging. Thus, managers do arelatively good job of tracking the S&P 500 Index.

39


Table 2: R2 Values of S&P 500 Index Mutual Funds

Panel A: R2 values of S&P 500 Index Mutual Funds.

Ticker R2

AASPX 100

AAFPX 100

ADIDX 100

ADIEX 100

WFSPX 100

BNSPX 100

SPFIX 100

CINIX 100

SBSDX 100

DSPCX 100

DSPNX 100

DSPIX 100

PEOPX 100

ETSPX 100

GIDIX 100

ISIIX 100

ISPIX 100

MDSRX 100

MASRX 100

SPIAX 100

SPIBX 100

SWPEX 100

SWPIX 100

SCPIX 100

KSAAX 100

SSPIX 100

TRQIX 100

SBSPX 100

SVSPX 100

SAPIX 100

PSPIX 100

40

Table 2: R2 Values of S&P 500 Index Mutual Funds

Panel A: R2 values of S&P 500 Index Mutual Funds.

Ticker R2


PWSPX 100

UASPX 100

UAIIX 100

USSPX 100

Panel B: Descriptive Statistics

Mean Standard Deviation Minimum Maximum

R2 100 0 100 100

This table presents the R2 values of S&P 500 Index mutual funds. Panel A presents the fund ticker and R2 foreach fund. The mean, standard deviation, minimum and maximum R2 values are displayed in Panel B.

RETURNS OF S&P 500 INDEX MUTUAL FUNDS

Elton, Gruber, and Busse argue that, in addition to tracking, investors should also beconcerned with an index fund’s returns relative to the index’s returns. Even if a fund has a hightracking measure, this may not translate into returns following the index. Following Elton, Gruberand Busse’s lead, I now analyze the returns on the S&P 500 Index mutual funds relative to the S&P500 Index over a four year time period. The data is provided by Morningstar and presented in Table3. Clearly, S&P 500 Index mutual funds underperform the S&P 500 Index. The average S&P 500Index mutual fund underperformed the S&P 500 Index by .6520% in 1999, .3928% in 2000, .4781%in 2001, and .3537% in 2002. Some funds underperformed the Index by as much as two percent.At least 92, 92, 97, and 92 percent of the funds underperformed the Index in 1999, 2000, 2001, and2002, respectively. Elton, Gruber, and Busse suggest that the performance of index funds wouldlikely differ from the returns on the index. Thus, my findings are not alarming.

Table 3: Returns of S&P 500 Index Mutual Funds Relative to the S&P 500 Index

Panel A: Deviation of Returns of S&P 500 Index Mutual Funds

Mean Stand Dev Minimum Maximum

1999 Deviation (%) -.6520 .4840 -2.0 0.1

2000 Deviation (%) -.3928 .4082 -1.4 0.6

2001 Deviation (%) -.4781 .3237 -1.7 0.1

2002 Deviation (%) -.3537 .3486 -1.2 0.4

41

Table 3: Returns of S&P 500 Index Mutual Funds Relative to the S&P 500 Index


Panel B: Number of Funds that Outperform and Underperform the S&P 500 Index

1999 2000 2001 2002

# that Outperform 1 2 1 3

(.0400) (.0714) (.0244) .0732)

# that had Same Performance 1 0 0 0

(0400) (.0000) (.0000) (.0000)

# that Underperform 23 26 40 38

(.9200) (.9286) (.9756) (.9268)

Number of funds in Sample 25 28 41 41

Panel A presents the deviation of returns of S&P 500 Index mutual funds relative to the Index over a four yearperiod. The mean, standard deviation, minimum and maximum of deviations of returns are presented. Panel Bpresents the number of S&P 500 Index Mutual Funds that outperform, underperform, and perform the same as theS&P 500 Index over a five year period. The percentages are in parentheses. Morningstar is the data source.

RELATIONSHIP BETWEEN THE RETURN AND THE EXPENSE RATIO

According to Collins (1999), assuming an index fund successfully replicates its index, thefund’s expense ratio may be the primary determinant as to whether the fund does better or worsethan comparable index funds. Elton, Gruber, and Busse also suggest that expenses may explain thedifferences in returns amongst index funds. They maintain that by purchasing and selling securities,a fund incurs transaction costs which result in returns below that of an index. The descriptivestatistics presented in Table 1 demonstrate that there are differences across funds in terms of theexpense ratio. The amount that a fund underperforms the S&P 500 Index varies across funds, anda portion of the differential performance may be explained by the difference in expenses betweenfunds.

In this section, S&P 500 Index mutual fund returns are regressed on their expenses todetermine whether the variation in returns across mutual funds is explained by the variation inexpenses. Lipper is the data source for this analysis since they rank S&P 500 Index mutual fundsbased upon expenses. A service that Lipper provides is that it ranks a fund based on minimizingexpenses relative to its peers for a three-year, five-year, and ten-year time period. Three, five, andten year return and expense statistics are presented in Table 4.

42


Table 4: Returns and Expenses of S&P 500 Index Mutual Funds

Mean Std Dev Min Max N

3 Year Returns -1.64 .425 -2.83 -1.01 151

3 Year Expenses 3.01 1.428 1 5 151

5 Year Returns -.650 .403 -1.66 -.03 100

5 Year Expenses 3.03 1.44 1 5 100

10 Year Returns 10.94 .247 10.42 11.42 37

10 Year Expenses 3.108 1.409 1 5 37

This table presents the mean, standard deviation, minimum, and maximum of three-year, five-year and ten-yearreturns and expenses. The number of observations (N) is also displayed. Lipper is the data source.

Since Lipper provides data over different time periods, several simple regression models canbe analyzed. Three-year, five-year and ten-year returns are regressed on the three-year, five-year,and ten-year expense scores, respectively. The results are presented in Table 5, Panel A. The simpleregression model is as follows:

RETi = ai + biEXPi + εi (1)

where RETi is the three-year return (Model 1), five-year return (Model 2), and ten-year return(Model 3) on mutual fund i, and EXPi is the three-year expense measure (Model 1), five-yearexpense measure (Model 2), and ten-year expense measure (Model 3) of fund i. Following theargument of Elton, Gruber, and Busse, the prediction is that the relationship is negative. Indeed, thisis what I find. The slope is -.1406 when employing three-year returns and expenses, -.1422 for five-year returns and expenses, and -.1395 for ten-year returns and expenses. All slope coefficients arestatistically significant at the 1% level. The test statistic is -6.5286 for the three-year data, -5.8612for the five-year data, and -7.7199 for the ten-year data. This is evidence that funds with lowerexpenses have higher returns. Also, this empirical evidence supports Collins’ contention that,assuming that the index fund successfully replicates its index (as I find), the fund’s expense ratiomay be the primary determinant as to whether the fund does better or worse than comparable indexfunds.

RELATIONSHIP BETWEEN THE RETURN AND THE AGE OF THE FUND

Replicating the S&P 500 Index is a very difficult task. If experience has any value, thenmanagers may get better at replicating the S&P 500 index over time. Fortin, et al, suggest that itwould be logical to assume that there is a direct correlation between fund performance and portfoliomanager experience. Following the argument of Fortin, et al, fund families with more experience

43


should have higher returns. There should be a positive relation between the experience of a fund,as measured by longevity, and its return. To determine if a relationship exists, returns are regressedon the age of the fund using Lipper data. The expectation is that the slope coefficient is positive.The simple regression model is as follows:

RETi = ai + biAGEi + εi (2)

where RETi is the return on mutual fund i, and AGEi is the age of fund i. As was mentionedpreviously, with Lipper providing data over different time periods, several simple regression modelscan be analyzed. The results are presented in Table 5, Panel B. The return is the three-year returnand the age is the 3, 5 or 10 years of existence of the fund in Model 1. In Model 2, the return is thefive-year return and the age is the 5 or 10 years in existence of the fund. I indeed observe a positive,highly significant relationship between the return and age of a fund. When regressing three-yearreturns on age, the slope coefficient is .0494 (test-statistic = 4.0889) and significant at the 1% level.The slope coefficient is .0479 (test-statistic = 3.0087) and significant at the 1% level whenregressing five-year returns on age. This is evidence that older, more established funds have higherreturns.

MULTIPLE REGRESSION

From the simple regression models above, I find that the expense measure and the age of thefund are highly correlated with the return on an S&P 500 Index mutual fund. In this section, Ianalyze whether or not any additional variation in returns can be explained by a multiple regressionmodel. The model is:

RETi = αi + β1EXPi +β2AGEi + εi (3)

where the dependent and independent variables were previously defined. The existence ofmulticollinearity may result in the expense ratio and/or the inception date not being significant ina multiple regression model. The results are presented in Table 5, Panel C. In Model 1, three-yearreturns are regressed on three-year expenses and age. In Model 2, five-year returns are regressedon five-year expenses and age. As presented, both independent variables remain highly statisticallysignificant with the same signs as from the simple regression models. The expense ratio isnegatively significant (one-percent level), while age is positively significant (one-percent level inModel 1, and five-percent level in Model 2) in both multiple regression models. Overall, bothmodels are highly significant (F-value of 30.6971 in Model 1, and 21.2512 in Model 2) and improvethe explanatory power relative to the simple regression models.

44


Table 5: Regression of Returns on Fund Characteristics

Panel A: Simple Regression Model: RETi = ai + biEXPi + εi

Indep Var Model 1 Model 2

3-yr Rets 5-yr Rets

Intercept -1.223 -.2191

(-17.04)*** (-2.69)***

EXP -.1406 -.1422

(-6.53)*** (-5.86)***

R2 .2224 .2595

N 151 100

Panel B: Simple Regression Model: RETi = ai + biAGEi + εi

Indep Var Model 1 Model 2

3-yr Rets 5-yr Rets

Intercept -1.925 -.9831

(-25.49)*** (-8.39)***

AGE.0494 .0479

(4.09)*** (3.01)***

R2 .1009 .0846

N 151 100

Panel C: Multiple Regression Model: RETi = α0 + β1EXPi +β2AGEi + εi,

Model 1 Model 2

Indep Var 3-yr Rets 5-yr Rets

Intercept-1.489 -.4943

(-15.37)*** (-3.65)***

EXP -.131 -.132

(-6.345)*** (-5.541)***

AGE.042 .035

(3.85)*** (2.51)**

R2 .2931 .3046

Adjusted R2 .2836 .2903

F-value 30.70*** 21.25***

N 151 100

45

Table 5: Regression of Returns on Fund Characteristics


**, *** indicate significance at the 5%, and 1% level, respectively.The fund characteristics are regressed on returns to determine whether the variation in returns across mutual fundsis explained by the variation in these fund characteristics. The dependent variable is RET. RET is the three-yearreturn, five-year return, and ten-year return. The independent variable is the expense measure (EXP) in Panel Aand the age of the fund (AGE) in Panel B. Simple regression models are presented in Panel A and Panel B, whilemultiple regression models are presented in Panel C. The regression coefficient is followed by the test statistic inparentheses.

PREDICTABILITY OF RETURNS

Elton, Gruber, and Busse note that investors should consider if they can predict the returnsthat a fund will generate in the future. An investor who is concerned about future returns, will likelystudy past returns when picking funds. Past returns might predict future returns. In this section, Iexamine if a statistically significant relationship exists between returns over time. Returns areregressed on lagged returns to determine whether the variation in returns across mutual funds isexplained by the variation in lagged returns. Since Morningstar and Lipper provide the returns onthe funds that they monitor, both data sources are employed in this analysis.

Using Lipper data, five-year returns are regressed on three-year returns (Model 1), ten-yearreturns are regressed on three-year returns (Model 2), and ten-year returns are regressed on five-yearreturns (Model 3). The results are presented in Table 6. I present evidence that performance ispositively correlated with past performance. The slope coefficient is positive and highly significant(all at the one percent level) for each of the three regression models.

Table 6: Regression of Returns on Lagged Returns

Model 1 Model 2 Model 3

Indep Var 5-yr Rets 10-yr Rets 10-yr Rets

Intercept.981 12.447 11.435

(34.78)*** (213.51)*** (543.53)***

3 Year Rets1.033 1.073

(59.509)*** (26.823)***

5 Year Rets1.043

(28.40)***

R2 .9730 .9510 .9561

N 100 39 39

46

Table 6: Regression of Returns on Lagged Returns


**, *** indicate significance at the 5%, and 1% level, respectively.Returns are regressed on lagged returns to determine whether the variation in returns across mutual funds isexplained by the variation in lagged returns. In Model 1, five-year returns are regressed on three-year returns. Model 2 presents the ten-year returns regressed on three-year returns. In Model 3, ten-year returns are regressedon five-year returns. The regression coefficient is followed by the test statistic in parentheses. Lipper is the datasource.

When employing Morningstar data, year 2000 returns are regressed on year 1999 returns,year 2001 returns are regressed on year 2000 returns, year 2002 returns are regressed on year 2001returns, and year 2003 returns are regressed on year 2002 returns. The results are presented in Table7, Panels A, B, C and D (Model 1). The slope coefficient is positive and highly significant (all atthe one percent level) for all four simple regression models. The slope (and test statistic) is .7425(9.1174), .5676 (7.4912), .7257 (5.9365), and .8646 (5.0414) when regressing 2000 returns on 1999returns, 2001 returns on 2000 returns, 2002 returns on 2001 returns, and 2003 returns on 2002returns, respectively.

Table 7: Regression of Returns at Time T+1, T+2, T+3, T+4 on Returns at Time T

SimpleRegressionModels:

Model 1: RET(T+1)i = αi + biRET(T)i + εi

Model 2: RET(T+2)i = ai + biRET(T)i + εi



Panel A: Independent Variable is 1999 Returns, Dependent Variables are 2000, 2001, 2002, 2003 Returns

Model 1 Model 2 Model 3 Model 4

Indep Var 2000 Rets 2001 Rets 2002 Rets 2003 Rets

Intercept-24.61 -23.15 -34.49 11.10

(-14.81)*** (-20.57)*** (-19.90)*** (6.58)***

1999 Returns.7425 .5308 .5913 .8304

(9.11)*** (9.63)*** (6.96)*** (10.04)***

R2.7832 .8011 .6781 .8144

N 25 25 25 25

Panel B: Independent Variable is 2000 Returns, Dependent Variables are 2001, 2002 and 2003 Returns

Model 1 Model 2 Model 3

Indep Var 2001 Rets 2002 Rets 2003 Rets

Intercept-6.929 -15.275 36.923

(-9.62)*** (-16.28)*** (39.11)***

47

Table 7: Regression of Returns at Time T+1, T+2, T+3, T+4 on Returns at Time T


2000 Returns.5676 .7557 .9366

(7.49)*** (7.65)*** (9.42)***

R2 .6833 .6926 .7736

N 28 28 28

Panel C: Independent Variable is 2001 Returns, Dependent Variables are 2002 and 2003 Returns

Model 1 Model 2

Indep Var 2002 Rets 2003 Rets

Intercept-13.472 43.417

(-8.90)*** (29.88)***

2001 Returns.7257 1.2507

(5.93)*** (10.65)***

R2 .4746 .7441

N 41 41

Panel D: Independent Variable is 2002 Returns, Dependent Variables is 2003 Returns

Model 1

Indep Var 2003 Rets

Intercept47.358

(12.30)***

2002 Returns.8646

(5.04)***

R2 .3945

N 41

*, **, *** indicate significance at the 10%, 5%, and 1% level, respectively.Returns are regressed on lagged returns to determine whether the variation in returns across mutual funds is explainedby the variation in lagged returns using Morningstar data. These are simple regression models. In Panel A, thereturns from 1999 are regressed on the returns from 2000, 2001, 2002, and 2003, respectively. The returns from 2000are regressed on the returns from 2001, 2002, and 2003, respectively, in Panel B. The returns from 2001 are regressedon the returns from 2002, and 2003, respectively, in Panel C. In Panel D, the returns from 2002 are regressed on thereturns from 2003. The regression coefficient is followed by the test statistic in parentheses.

I now examine the association between returns and past returns lagged for two, three andfour periods as presented in Table 7. In Panel A, returns from 2001, 2002, and 2003, respectively,are regressed on returns from 1999. Panel B presents the results of returns from 2002 and 2003,respectively, regressed on returns from 2000. Returns from 2003 being regressed on returns from2001 are displayed in Panel C. The simple regression models are:

48


RET(T+2)i = ai + biRET(T)i + εi (Model 2) (4)



where: RET(T+2)i is the return on mutual fund i two time periods subsequent to time T; RET(T+3)I

is the return on mutual fund i three time periods subsequent to time T; RET(T+4)i is the return onmutual fund i four time periods subsequent to time T; and RET(T)i is the return on mutual fund i attime period T. When examining the association between returns and returns lagged two, three, andfour periods, I get very strong results. As presented in Table 7, the slope coefficient is positive andsignificant at the one percent level for every simple regression. When 2001, 2002, and 2003 returns,respectively, are regressed on 1999 returns, the slope coefficient (and test statistic) is .5308 (9.6256),.5913 (6.9612), .8304 (10.0461), correspondingly. The slope (and test statistic) is .7557 (7.6545)and .9366 (9.4278) when 2002 and 2003 returns, respectively, are regressed on 2000 returns. Theslope is 1.2507 with a test statistic of 10.6509 when returns from 2003 are regressed on returns from2001. Thus, I present additional evidence that returns are positively correlated with past returns.

CONCLUSIONS

In this study, I analyze the performance of S&P 500 Index mutual funds. The great appealof researching Index funds is that there are limited opportunities in the equity markets to studyhomogeneous investment vehicles. I investigate the issues that Elton, Gruber, and Busse contendthat index fund investors should care about: returns and tracking. This study differs from previouspapers in a number of ways. Unlike other researchers who use Morningstar exclusively as their datasource, I employee two popular mutual fund data sources to conduct my analysis, Morningstar andLipper. Also, I analyze both tracking and performance, and examine possible influences onperformance.

My results reveal that managers do a relatively good job of tracking the S&P 500 Index.However, despite the high tracking measures, S&P 500 Index funds clearly underperform the S&P500 Index. I present evidence that funds with lower expenses have higher returns and older, moreestablished funds have higher returns. I also find that returns are positively correlated with pastreturns over several time periods. With so many index funds to analyze, investors may simply beoverwhelmed with information. This paper provides research to simplify the analysis for investors.

49


REFERENCES

Blume, M., & R. Edelen (2004). S&P 500 Indexers, Tracking Error, and Liquidity. The Journal of PortfolioManagement, Spring, pp. 37-46.

Carhart, M. (1997). On Persistence in Mutual Fund Performance. Journal of Finance, 52, pp. 57-82.

Chen, H., G. Noronha, & V. Singal (2005). Index Changes and Unexpected Losses to Investors in S&P 500 and Russell2000 Index Funds, Working Paper.

Collins, P. (1999). Monitoring Passively Managed Mutual Funds. Journal of Investing, 8, p 49-61.

Cresson, John E. (2002). R2: A Market-Based Measure of Portfolio and Mutual Fund Diversification. QuarterlyJournal of Business and Economics, 41, Summer/Autumn, pp. 115 - 143.

Elton, E., M. Gruber, & C. Blake (1996). The Persistence of Risk-Adjusted Mutual Fund Performance. Journal ofBusiness, 69, pp. 133-157.

Elton, E., M. Gruber, & J. Busse (2004). Are Investors Rational? Choices among Index Funds. Journal of Finance, 59,pp. 261-288.

Fortin, R., S. Michelson, & J. Jordon-Wagner (1999). Does Mutual Fund Manager Tenure Matter? Journal of FinancialPlanning, 12, pp. 72-79.

Frino, A., & D. Gallagher (2001). Tracking S&P500 Index Funds. Journal of Portfolio Management, 28, pp. 44-55.

Frino, A., D. Gallagher, A. Neubert, & T. Oetomo (2004). Index Design and Implications for Index Tracking: Evidencefrom S&P 500 Index funds. Journal of Portfolio Management, Winter, pp. 89-95.

Goetzmann, W., & M. Massa (2002). Daily Momentum and Contrarian Behavior of Index Fund Investors. Journal ofFinancial and Quantitative Analysis, 37, pp. 375-389.

Grinblatt, M. & S. Titman (1989). Mutual Fund Performance: An Analysis of Quarterly Portfolio Holdings. Journalof Business, 62, pp. 393-416.

Gruber, M. (1996). Another Puzzle: The Growth in Actively Managed Mutual Funds. Journal of Finance, 55, pp. 783-810.

Keim, D. (1999). An Analysis of Mutual Fund Design: The Case of Investing in Small-Cap Stocks. Journal of FinancialEconomics, 51, pp. 50-69.

Malkiel, B. (1995). Returns from Investing in Equity Mutual Funds from 1971 to 1991. Journal of Finance, 50, pp.549-572.

Perold, A. (1988). The Implementation Shortfall: Paper Versus Reality. Journal of Portfolio Management, 20, pp. 4-9.

50


51


2008-2009 TUITION AND FEES IN HIGHEREDUCATION

Brian T. Kench, The University of TampaH. Scott Wallace, University of Wisconsin-Stevens PointKevin Neuman, University of Wisconsin – Stevens Point

ABSTRACT

Using data set from U.S. News & World Report, the Princeton Review, and the Associationto Advance Collegiate Schools of Business, this paper isolates 9 choice variables, 5 controlvariables, and 18 rankings variables that account for between 79 and 90% of the variation in tuitionand fees across 518 institutions of higher learning in the United States. It is hoped that the resultswill give guidance to schools by quantifying the costs and benefits of making a given change to theirtuition and fee structure.

INTRODUCTION

The degree of heterogeneity among institutions of higher learning in the U.S. is considerable.Universities and colleges are differentiated across a wide variety of attributes: public versus private,graduate programs versus no graduate programs, and religious versus secular to name a few. Inaddition, many colleges and universities focus their resources in excelling in specialized fields. Thisdiversity is a great strength and is one of the reasons why the United States remains the world leaderin education at the university level.

The attributes colleges and universities possess obviously are critical in determining thechoices of attendees. Yet, administrative decisions regarding the level of tuition and fees often arehaphazard and unrelated to the strengths of the university. In an era of declining taxpayer supportand unfavorable demographic trends, public institutions, in particular, need to set tuition to thespecific benefits that they offer students. To do this wisely, schools must assess their competitivepositions vis-à-vis other institutions. Using data from U.S. News & World Report, supplemented bythe Princeton Review and the Association to Advance Collegiate Schools of Business, we isolate9 choice variables, 5 control variables, and 18 rankings variables that account for 79 to 90 percentof the variation in tuition and fees across 518 institutions of higher learning. The objective of thisanalysis is to analyze market outcomes econometrically in an effort to assist administrators in theirtask of setting tuition and fees given their school’s market attributes.

52


Mapping Competitive Markets

The theoretical foundation that underpins the analysis used in this paper dates back to the1960s with the work of economist Lancaster (1966, 1971). Lancaster introduced a new approach toconsumer theory that emphasized the importance of product characteristics in the utility maximizingchoices of consumers. This characteristics approach “assumes that consumer demand is directed nottowards products per se but rather towards product characteristics. For example, when searching fora computer, consumers look for microprocessor speed, RAM capacity, hard disk capacity, screensize, and so forth. A consumer’s valuation for a particular computer is the sum of the valuations foreach particular characteristic” (Cabral, 2000, 207). The characteristics approach has led to effortsin developing empirical methods for hedonic models, which are used to estimate the demand forconsumer products. Drawing upon data on car models sold in the U.S. from 1971 to 1990,economists, for example, have estimated “own- and cross-price elasticities as well as elasticities ofdemand with respect to vehicle attributes (such as weight or fuel efficiency)” (Berry, Levinsohn &Pakes 1995, 841). The value of this analysis is not limited to the economics discipline. For example,Gorman (1980) analyzed the egg market. Petrin (2002) studied the minivan market. And Benkardand Bajari (2004) estimated personal computer demand.

D’ Aveni (2007) adapted the characteristics approach to create what he calls a “price-benefitpositioning map.” A price-benefit positioning map is similar to Lancaster’s approach because itshows the relationship between the primary benefit that a product provides to customers and theprices of all the products in a given market (D’Aveni 2007, 112). The author outlines a number ofsteps that need to be completed in creating a price-benefit positioning map. First, the firm mustdefine the relevant market for its product. This requires that it identify other products that are similarto the firm’s own product in the sense that they satisfy the same consumer needs. In addition, firmsmust define the geographic region within which competition occurs. Second, firms need to collectdata on prices of these products and on the benefits that these products provide to consumers.Drawing on this data, firms then perform regression analysis to find the relationship between productprice (the dependent variable) and product benefits (the independent variables). Third, try to isolatethe primary benefit (or group of correlated benefits) that accounts for the greatest variation in pricesamong competing products (D’Aveni, 2007).

Building on the above analysis, this paper seeks to isolate the variables that explain thevariation in tuition and fees across institutions of higher education in the U.S. The results, we hope,will give guidance to enrolment management personal at colleges and universities by quantifyingthe costs and benefits of making a given change to their tuition and fee structure.

53


EMPIRICAL MODEL

The empirical results from the framework outlined by D’Aveni (2007) come from a basicregression analysis. We estimate the following model to investigate the effects of various institutioncharacteristics on tuition:

Tuition & fees = a + bi*choice variables+ bj*control variables + by*rankings + e.

The dependent variable is the undergraduate tuition and fees for the 2008-2009 academicyear. We estimate two OLS regressions reporting the results later in the paper. The first model usesin-state tuition and fees for public institutions and the second uses out-of-state tuition and fees forpublic institutions. The tuition and fees amount for private institutions is the same in both models.As the majority of students pay in-state tuition we consider the in-state model to be the baselineestimates. However, a number of students do pay out-of-state rates making this pricing decisionimportant as well. Differences between students who pay in-state versus out-of-state tuition may alsocause differences in the relevant factors for each market, information which can be useful foradministrators in charge of setting tuition rates.

We divide the explanatory variables of the model into choice, control, and ranking groups.The primary variables in the model are the choice variables, which are defined as variables that aninstitution has the freedom to change at their discretion. While tuition may vary for other reasons,if the variables are out of the control of the institutions they are not variables which can bemanipulated to strategically price tuition, thus making them relatively less important. The controlvariables are variables which likely have an effect on tuition, but are not characteristics which canbe controlled by the institution. We include these variables to avoid biasing the choice variablecoefficients, but also to help separate possible submarkets of higher education due to region andschool type. Finally, we include a series of ranking variables constructed by various third parties.The ranking variables are included to see if outside perceptions of the institution affect pricingdecisions.

Data

The next step of the theoretical framework is to define the appropriate market. Given ourbackground and the prominent place of the United States in higher education, we use the U.S. as therelevant geographic market. Even within this geographic market, prospective students have theopportunity to consider a variety of colleges and universities. Thus, we include 267 master’s leveluniversities, 133 national universities, and 118 liberal arts colleges as described in the 2009 U.S.News & World Report (USNWR). Fourteen additional institutions had incomplete USNWR dataand have been dropped from the study. Thus, the dataset used here contains 518 observations, which

54


represents 23 percent of all colleges and universities in the U.S. (Although we prefer an analysisof the total population, limited data have constrained our analysis to the available sub-group. Giventhe distribution of the sample the results should be representative of the entire population.)

The USNWR data set contains a variety of information about institutions, such as tuition,student to faculty ratio, and mean graduation rate. USNWR also compiles rankings of institutionswithin regions and school type. While gathered by a third party not actively involved in highereducation, the data and rankings grab the attention of both sides of the education decision.Administrators actively advertise their position in the rankings lists, while prospective students usethe data to compare institutions, trusting that the data is accurate and comparable. As such theUSNWR data is an ideal source of information for our efforts to map the relationship between tuitionand the characteristics valued by prospective students.

We also draw information from a few other sources to supplement our analysis. Toincorporate other elements of the college experience we include data from 12 of the most commonPrinceton Review (PR) ranking categories for 2009. While primarily known for their test preparationservices Princeton Review also compiles top ten lists of institutions with regard to academic issues,but also with regard to broader factors such as politics, quality of life, and social scene. To test forthe effects of business school accreditation we include data on membership status during the fall of2008 in the Association to Advance Collegiate Schools of Business (AACSB). The mission ofAACSB is to advance quality management education worldwide through accreditation and throughleadership. Having the AACSB accreditation is thought to be a signal of educational excellence.Numerous other accreditation bodies exist which could be included in this type of analysis, both forbusiness schools and other disciplines. However, AACSB is the most relevant accreditation to thediscipline of the current analysis and is the only one which is included.

Choice variables

The model has nine choice variables which capture dimensions institutions can change ifthey desire. The first two variables capture the highest degree awarded. USNWR categorizesinstitutions as national, masters, or liberal arts, which turn out to be proxies for the highest degreeawarded. At all national universities the highest degree awarded is a PhD, at all masters universitiesthe highest degree awarded is a masters, and at all liberal arts colleges the highest degree awardedis a bachelors. The masters variable has institutions that award masters level degrees as the highestdegree coded 1, while all others are coded 0. The PhD variable has institutions that award PhD leveldegrees as the highest degree coded 1, while all others are coded 0. Both categories are measuredrelative to institutions where a bachelors degree is the highest degree awarded. Institutions have theability to change the highest degree that they award to define the markets in which they compete.The variables included should capture how tuition varies across the degree types. The results shouldinform other institutions looking to enter a different market by showing whether tuition can be

55


changed to recoup the costs of a higher degree. The next two choice variables measure the percentof classes with less than 20 students and the percent of classes with greater than 50 students. Thesevariables are included to capture the effect of relatively small and large classes, both of which maybe manipulated to attract students or to cut costs. Smaller class sizes are a draw for some studentsbut may come at a price for the institution if they need to hire faculty or cut enrolment to achievethem. On the other hand, larger classes could be a sign of trying to get more production from thesame faculty and may allow institutions to lower tuition as a result.

The fifth choice variable measures the students-to-faculty ratio. This commonly citedvariable is a rough measure of the intensity of use of the faculty inputs in the education process.Once again, low student-to-faculty ratios may be attractive to students allowing institutions to chargehigher tuition, but they also may be more costly forcing institutions to raise tuition. The sixth choicevariable measures the average graduation rate. USNWR defines the average graduation rate as thepercentage of entering freshmen who graduated within a six-year period or less, averaged over theclasses entering from 1999 through 2002. This calculation excludes students who transferred intothe school. If a school submits fewer than four years of graduation rate data, then the average isbased on the number of years that are submitted. As graduation is the end product of the educationconsumption decision, the ability to bring students to graduation is an important measure of theefficiency of the institution. The next two variables capture the size of the student population andare included to check for economies of scale in the provision of education. The variables also mightreflect different willingness to pay of students for different size institutions. Two dummy variableshave been created for this purpose. Institutions that have an undergraduate student populationbetween 6,000 and 8,999 are coded 1, while all others are coded 0. Institutions that have a studentpopulation of 9,000 and higher are coded 1, while all others are coded 0. Both categories aremeasured relative to institutions that have a student population between 1 and 5,999. The finalchoice variable captures whether or not an institution has the AACSB business school accreditation.Institutions that are AACSB accredited are coded 1, while those that are not are coded 0 (eitherbecause their business school is not accredited or because they do not have a business school). Asachieving AACSB, or any accreditation, is a costly but potentially marketable endeavor, we includethis variable to see if tuition changes as a result.

Control Variables

The model has five control variables which are included to capture variations in tuition dueto characteristics out of the control of the institution. The first of these variables controls forreligious affiliation: secular institutions, with no religious affiliation, are coded 0, while institutionsthat have a religious affiliation are coded 1. The variable should capture whether students are willingto pay a premium to attend a religious institution. The second variable controls for privateinstitutions: public institutions are coded 0, while private institutions are coded 1. Once again, this

56


variable can capture different willingness to pay for private and public institutions, but can also bethought to measure the size of any subsidy received by public institutions. The last three controlvariables are for the institution’s region of the country. USNWR divides the United States into fourregions: north, south, west, and mid-west. Here we have coded three dummy variables usingregional boundaries defined by USNWR. Institutions located in the south (west, mid-west) are coded1, while institutions outside of the south (west, mid-west) are coded 0. The three regional dummyvariables are measured relative to institutions located in the north. The three regional controls shouldaccount for cost of living differences in each region, but also may measure different willingness topay for education in the regions.

Rankings Variables

Lastly, the model has two sources of ranking variables: six variables from USNWR and 12variables from the PR. The six USNWR ranking dummy variables are top 10 masters-north, top 10masters-south, top 10 masters-west, top 10 masters-mid-west, top 10 national, and top 10 liberalarts. Institutions that are ranked in the top 10 in each USNWR ranking category are coded 1, whileall other institutions are coded 0. USNWR rankings are included to measure whether the third partyrankings allow institutions to raise tuition due to the recognition.

The 12 Princeton Review ranking dummy variables are top 10 institutions with the bestclassroom experience, top 10 institutions with the best Greek life, top 10 institutions with the leasthappy students, top 10 institutions with the happiest students, top 10 institutions where students arehappiest with financial aid, top 10 institutions that are socially liberal, top 10 socially conservativeinstitutions, top 10 institutions with the most diverse student population, top 10 institutions with themost homogeneous student population, top 10 biggest party schools, top 10 institutions wherestudents rarely study, and top 10 institutions where students always study. Institutions that areranked in the top 10 in each PR ranking category are coded 1, while all other institutions are coded0. The PR rankings include a much more diverse set of institution characteristics, some of whichadministrators may not actually desire. However, the various social characteristics may be valuedby prospective students and therefore may be characteristics than can be exploited with highertuition.

RESULTS

We estimate the model using in-state tuition and fees as our baseline model, presenting theresults in Table 1. Turning to the choice variables we find a number of significant results whichinstitutions might be able to manipulate for their advantage. First, relative to the 118 institutions thatonly award a bachelors degree, tuition and fees at the 133 institutions where the highest degreeawarded is a PhD are lower by $3,050.14, whereas, tuition and fees are lower by $3,920.69 at the

57


267 institutions where the highest degree awarded is a masters, ceteris paribus. There are at least twopossible explanations for this difference. First, advanced degree programs typically charge highertuition and fees, and thus some pressure to raise undergraduate tuition and fees is diminished. Thesecond, and the more pessimist perspective, is that schools that offer advanced degree programsspread themselves too thin. A variety of resources are funneled into these advanced degree programsat the expense of the undergraduate degree program, and the effect of this diminution is measuredby the market in the form of lower tuition and fees for institutions in this category. Administratorsare often uncertain of the opportunity costs of their actions. In many cases administrators only countthe benefits, e.g., tuition and fees from advanced degree programs. However, the coefficients on thevariables Masters and PhD offer a market measure of the opportunity cost of having advanceddegree programs.

The variables percent of classes with less than 20 students and percent of classes with morethan 50 students capture the marginal effects of changing the number of seats offered in each class.We find that while the small class size variable is insignificant, relatively large classes do have asignificant effect on tuition and fees. The coefficient of the percent of classes with greater than 50students reveals that if the percent of classes with greater than 50 students decreases by onepercentage point, tuition and fees would increase by $113.26 per student. This result suggests thatlowering the percent of classes with 50 or more students generates a bigger bang-for-the-buck thanincreasing the percent of classes with less than 20 students. While both moves would likely raisecosts due to the necessity of hiring more instructors, breaking up larger sections into average sizeclasses is a more marketable move. These results make the monetary benefits associated withaltering class size clearer, and would seem to refute the idea that prospective students value smallclass sizes.

The students-to-faculty ratio is also commonly advertised in an effort to attract prospectivestudents. Of the 518 institutions in our dataset, the mean students-to-faculty ratio is 13.4. Themarginal effect of decreasing this ratio by one percentage point is $287.97 per student, ceterisparibus. To put this number in perspective, for an average sized institution of 9,157 students, withthe mean students-to-faculty ratio, hiring 10 additional faculty members would reduce the students-to-faculty ratio by 0.2 percentage points. This decrease in the ratio would raise tuition by $57.59 perstudent on average, for an overall increase in tuition receipts of over $527,000. This result suggeststhat unless the institution can hire 10 faculty members for less than $527,000, which may be quitedifficult, it may not be cost effective to compete through this mechanism.

The mean graduation rate at the 518 institutions in our dataset is 67.8, which informs us that67.8 percent of all freshman graduate within 6 years. For each incremental increase in the graduationrate above the mean, institutions charge $177.87 more in tuition and fees, ceteris paribus. Studentsare willing to pay more for higher than mean graduation rates and less for lower than meangraduation rates. While there are likely a variety of ways institutions can work to improve their

58


graduation rate, the result gives an idea of the monetary benefit associated with the improvementand can be used to evaluate the cost effectiveness of various techniques.

Our dataset has a mean institution size of 9,157 students and contains 301 small institutionswith a student population below 5,999, 58 medium sized institutions with a student populationbetween 6000 and 8,999, and 159 large institutions with a student population greater than 9,000.Relative to small institutions, medium sized institutions charge $1,960.87 more per student, whilelarge institutions charge $2,119.55 more per student, ceteris paribus. The results may reflectdiseconomies of scale in the provision of education, but also may reflect a greater willingness to payfor larger institutions with the amenities they offer.

Finally, in our dataset, 248 institutions are accredited by AACSB. Tuition and fees atinstitutions with the AACSB accreditation is $1,036.66 higher relative to non-AACSB accreditedinstitutions, ceteris paribus. For an average sized institution the coefficient can give administratorsa monetary benchmark to weigh the costs of achieving the accreditation. Some caution should beexercised when interpreting the results, however. While the AACSB variable likely reflects the valueof an accredited business program, if institutions with AACSB accreditation are also more likely towork towards other accreditations, the variable may be picking up the value of other accreditationsas well.

With regard to the control variables, tuition and fees at the 206 religiously affiliatedinstitutions is lower by $2,577.78, ceteris paribus. This result may reflect a lower willingness to payfor religious institutions, or perhaps a greater ability to raise donor money due to their religiousaffiliation. Tuition and fees at the 359 private institutions is higher by $19,727.95. Stated differently,the average federal, state, and local subsidy per student at public institutions in the 2008-09academic year was $19,727.95 per student. This per-student public subsidy has increased by$2,497.98 when compared to a similar analysis made in Kench and Wallace (2010).

To measure regional differences in tuition and fees we used USNWR’s regionalcategorization, which resulted in 179 northern, 104 southern, 107 western, and 128 mid-westerninstitutions. Relative to institutions in the north, tuition and fees at institutions in the south are lowerby $3,227.98, in the west are lower by $1,142.01, and in the mid-west are lower by $2,013.51,ceteris paribus. It is interesting to observe the deep discount for southern institutions – 2.8 times thatof institutions in the west and 1.6 times that of institutions in the mid-west. This difference reflectsregional cost of living differences, but also likely reflects a different willingness to pay across theregions. While administrators cannot easily change their region of the country, they should useinstitutions in their region as comparisons when setting tuition.

Checking for effects of outside perceptions, we find the rankings data from USNWR and PRhave mixed results. First, consider the USNWR ranking data. After controlling for the otherindependent variables in our model, only the variable USNWR top 10 masters west is significant.Institutions ranked in the top 10 in the masters west region receive a premium of $2,548.01 perstudent, ceteris paribus. This result suggests there is some benefit to appearing in the rankings,

59


although it does not seem to be widespread. Second, consider the PR rankings. Interestingly, the top10 institutions where students are happiest with financial aid actually have tuition and fees that arelower by $4,292.81, perhaps reflecting a lesser need for financial aid given the lower tuition. Inaddition, the top 10 most socially conservative institutions have tuition and fees that are lower by$4,334.84 per student. This reduced tuition could possibly reflect fewer available amenities due toa lower demand by conservative students. However, students at the top 10 most homogeneousinstitutions are willing to pay $2,529.97 more per student for a homogeneous environment, ceterisparibus.

The results in Table 1 using in-state tuition rates likely reflect the primary decision forinstitutions as the majority of students pay in-state rates. However, a non-trivial number of studentsdo pay out-of-state tuition, and may base their decisions on different characteristics. To mostaccurately price this market administrators should take any differences into consideration. Table 2reports the regression results when the dependant variable for public institutions contains out of statetuition and fees. In general the results in Table 2 are quite similar to the baseline results using in-state tuition. The choice variables masters, PhD, students to faculty ratio, mean graduation rate,student population between 6000 and 8999; the control variable religious, private, south, mid-west;and the ranking variables PR top 10 institutions where students are happiest with financial aid, andPR top 10 socially conservative institutions all remain statistically significant with the same sign –although the magnitude of each coefficient has changed somewhat.

Table 1: Full Regression Estimates, Using In-State Tuition & Fees

Parameter Estimate p-value

(Constant) 4285.52 .11

Bachelors 0

Masters -3920.69 <.00

PhD -3050.14 <.00

Percent of classes with less than 20 students 10.97 .54

Percent of classes with more than 50 students -113.26 .05

Students to faculty ratio -287.97 .00

Mean graduation rate 177.87 <.00

Student population between 1 and 5,999 0

Student population between 6000 and 8999 1960.87 .00

Student population over 9000 2119.55 .00

No AACSB 0

AACSB 1036.66 .04

Secular 0

60

Table 1: Full Regression Estimates, Using In-State Tuition & Fees



Religious -2577.78 <.00

Public 0

Private 19727.95 <.00

North 0

South -3227.98 <.00

West -1142.01 .03

Mid-west -2013.51 <.00

USNWR top 10 masters north 260.04 .85

USNWR top 10 masters south 1817.75 .16

USNWR top 10 masters west 2548.01 .05

USNWR top 10 masters mid-west -1233.82 .34

USNWR top 10 national universities -95.20 .95

USNWR top 10 liberal arts 248.66 .86

PR top 10 institutions with the best classroom experience 1059.67 .48

PR top 10 institutions with the best Greek life -81.45 .95

PR top 10 institutions with the least happy students -758.55 .70

PR top 10 institutions with the happiest students 1094.74 .43

PR top 10 institutions where students are happiest with financial aid -4292.81 .01

PR top 10 institutions that are socially liberal 2501.84 .10

PR top 10 socially conservative institutions -4334.84 .01

PR top 10 institutions with the most diverse student population -1137.95 .40

PR top 10 institutions with the most homogeneous student population 2529.97 .08

PR top 10 biggest party schools -783.48 .61

PR top 10 institutions where students rarely study -635.56 .71

PR top 10 institutions where students always study -13.31 .99

Adjusted R-square .901; F-statistic = 147.91, p-value = <.00; no evidence of multicollinearity because noindividual variable inflation factor (VIF) is greater than 10.

We do find a few differences in the results, however. The choice variables percent of classeswith greater than 50 students, student population over 9000, and AACSB are not significant in theout-of-state model suggesting they are not factors in determining prices in this market. Perhaps themost interesting result is that for AACSB which suggests that students willing to move to another

61


state for their education are not willing to pay more for the distinction. In addition, the estimate ofthe control variable west is not significant. Lastly, the rankings variables USNWR masters west andPR top 10 institutions with the most homogeneous student population lose significance, while twoof the PR rankings variables, PR top 10 institutions that are socially liberal and PR top 10institutions with the most diverse student population, gain marginal significance in the out-of-statemodel. While somewhat interesting, the marginal significance of the rankings results in both Table1 and Table 2 suggests the difference in the models is not extreme.

Table 2: Full Regression Estimates, Using Out-of-State Tuition & Fees


(Constant) 7671.64 .01

Bachelors 0

Masters -3474.57 <.00

PhD -2035.09 .02

Percent of classes with less than 20 students 24.89 .19

Percent of classes with more than 50 students 37.44 .53

Students to faculty ratio -272.57 .00

Mean graduation rate 243.62 <.00

Student population between 1 and 5,999 0

Student population between 6000 and 8999 1537.76 .03

Student population over 9000 -487.79 .74

No AACSB 0

AACSB 636.82 .22

Secular 0

Religious -2194.25 <.00

Public 0

Private 9536.04 <.00

North 0

South -1781.42 .00

West 122.12 .83

Mid-west -1678.05 .00

USNWR top 10 masters north 687.21 .64

USNWR top 10 masters south 2087.66 .12

USNWR top 10 masters west 2035.07 .13

USNWR top 10 masters mid-west -1521.21 .26

62

Table 2: Full Regression Estimates, Using Out-of-State Tuition & Fees



USNWR top 10 national universities -2072.87 .21

USNWR top 10 liberal arts -487.79 .74

PR top 10 institutions with the best classroom experience 596.87 .70

PR top 10 institutions with the best Greek life -1060.24 .46

PR top 10 institutions with the least happy students 913.77 .66

PR top 10 institutions with the happiest students 223.90 .88

PR top 10 institutions where students are happiest with financial aid -4267.42 .01

PR top 10 institutions that are socially liberal 2702.04 .09

PR top 10 socially conservative institutions -6520.56 <.00

PR top 10 institutions with the most diverse student population -2592.38 .07

PR top 10 institutions with the most homogeneous student population 2494.02 .10

PR top 10 biggest party schools 776.95 .62

PR top 10 institutions where students rarely study -145.91 .93

PR top 10 institutions where students always study -318.92 .85

Adjusted R-square .791; F-statistic = 62.32, p-value = <.00; no evidence of multicollinearity because no individualvariable inflation factor (VIF) is greater than 10.

What we have here are end points for the value of product characteristics in higher education.Because some public students do pay out of state tuition and fees, they certainly see the worldthrough the lens of Table 2. However, many more students pay in state tuition and fees and thus seethe higher education landscape though the lens of Table 1. It must also be noted that once thedecision to attend a particular institution is made, some out of state students are able to capture instate prices by documenting relevant information. This effect places more weight on the resultpresented in Table 1; however, the results presented in Table 2 cannot, be dismissed.

SUMMARY AND CONCLUSIONS

It is hoped that administrators at colleges and universities around the U.S. will use the resultsof our analysis as a road map to help guide them in pricing their product optimally. The resultsoffered here give important clues as to how the market would respond to decisions that enrolmentmanagers might make. For example, the data reveal that students do not value small classes, but arewilling to pay more to lower the percent of classes with greater than 50 students. If budgets are tight,and we know that they are, then institutions should focus on decreasing the number of classes withmore than 50 students before they increase the number of classes with less than 20 students. We also

63


find that lower student-to-faculty ratios and higher graduation rates are valued on the market withstudents willing to pay higher tuition. The empirical results we obtain give administrators the toolsto test whether these are cost effective ways of competing in the market. If the costs of new facultyand efforts to raise graduation rates are too high, the additional tuition might not be enough to makethe moves profitable.

In addition, if your institution has a business school, it might be worth the enormous effortof gaining AACSB accreditation. The data suggest that by adopting AACSB’s standards for aquality management education, the institution as a whole seems to gain a positive externality.Administrators are often curious as to whether gaining AACSB accreditation is worth the significanthurdles an institution must go over to do so. Our empirical evidence identifies a benefit associatedwith having AACSB accreditation that can be compared with the costs incurred throughout theprocess.

Institutions where a graduate program exists need to take note. Something interesting ishappening in the market such that undergraduate students are paying less if a graduate programexists, relative to undergraduate students at institutions with no graduate programs. If this differenceis not due to revenue shifting across degree types, but rather a declining quality of the undergraduateexperience, administrators should be concerned about entering this segment of the marketplacebecause the aggregate behavior of consumers is certainly signaling that something interesting isgoing on.

Finally, for all the conversation about campus Greek life, top party schools, and other PRcategories, we observe, in most cases, no significant difference in the tuition and fees charged bythe top ten institutions in these categories and all others. The same can be said for the USNWRacademic rankings. These results suggest that putting too much stock in the outside opinions of theinstitution is unwarranted, as market participants do not appear to be willing to pay for thedistinctions.

Although our study produces some interesting results, it does have a few limitations. Whileour results show in which direction tuition and fees move given various institution characteristics,we are not able to necessarily identify why the tuition is changing. For example, increases in tuitioncould be due to greater willingness to pay for the characteristic by students, but also may be due tonecessary price increases to offset higher costs. Future research should attempt to identify themechanism behind the tuition changes. It would also be informative to include other accreditationdistinctions in the analysis to see if it is AACSB accreditation specifically which is valued on themarket, or whether the variable is capturing a recognition of higher quality instruction overall. Giventhese issues we hope our analysis provides a useful starting point to aid and inform administrators.

REFERENCES

Benkard, C.L., and Bajari, P. (2004). Demand Estimation with Heterogeneous Consumers and Unobserved ProductCharacteristics: A Hedonic Approach, NBER working paper 10278.

64


Berry. S., Levinsohn, J., and Pakes. A. (1995). Automobile Prices in Market Equilibrium, Econometrica, Vol. 63, pp.841-890.

Cabral, L. (2000). Introduction to Industrial Organization. Cambridge: The MIT Press.

D’ Aveni, R. (2007, November). Mapping Your Competitive Position, Harvard Business Review, November, pp. 110-120.

Gorman, T. (1980). A Possible Procedure for Analyzing Quality Differentials in the Egg Market, Review of EconomicStudies, 47(5): pp. 843-856.

Kench, B. T., and H. Scott Wallace. (2010). Mapping 2007-08 Tuition and Fees in Higher Education, Journal of AppliedBusiness Research, 26(1): pp. 11-18.

Lancaster, K. (1966). “A New Approach to Consumer Theory,” Journal of Political Economy, Vol. 74, No. 2, April, pp.132-157.

Lancaster, K. (1971). Consumer Demand. New York: Columbia University Press.

Petrin, A. (2002). Quantifying the Benefits of New Products: The Case of the Minivan, Journal of Political Economy,100: pp. 705-729.

65


CONVERGENCE OF E-TAILING ANDSOCIAL NETWORKING

Maria Luisa C Delayco, Hankuk University of Foreign StudiesBrandon Walcutt, Hankuk University of Foreign Studies

ABSTRACT

Social networks are fundamentally changing business models and shaping markets as theyempower internet retailers or e-tailers, potential and current consumers and consumer groups. E-tailing is growing, allowing small and medium e-tailers (SMETs) to offer varied products to themarket. Being a SMET once posed structural and functional limitations, but incorporating thecapabilities of the web and the presence of social networks allows these firms to create mechanismsand strategies for marketing, communication and distribution that level the playing field. Two casesare examined to extract internet retailing strategies used by SMETs that particularly make use ofsocial networks in their communication strategies. Key issues for customers and e-tailers arediscussed. Critical success factors are identified and strategies are suggested for SMETs.

This paper is unded by Hankuk University of Foreign Studies, Seoul, Korea.

INTRODUCTION

Social networks and their influence on internet retailers or e-tailers are becoming more andmore evident in today’s virtual world (Enders & Jelassi, 2000). Although the true effects of socialnetworks cannot yet be measured, a growing number of companies are beginning to incorporate thischannel into their marketing and communications strategies. Even large companies like DELL,Kodak and General Motors are using tools, such as Twitter, to buff their brands and providecustomer service (King, 2008).

Social Networking Sites (SNS) are recognizing this development and are providing facilitiesand services for e-tailer use. Notable examples of these services are Facebook Connect andBusiness on Twitter. These services are free and are making themselves attractive to small andmedium sized firms by helping reduce e-tailing marketing costs.

The usage of social networking by e-tailers is also expected to grow as the online marketgains ground. Studies indicate that digitized products (software, music, books etc.) are expected tomake a radical shift into e-tailing (Enders & Jelassi, 2000; Grewal, Iyer & Levy, 2004). This shiftcan also be expected for market leaders of touch and feel products (clothes, grocery items etc.)

66


because of their prior customer base, but not for small players (Grewal, Iyer & Levy, 2004). Thesesmaller players have found it difficult to compete in the e-tailing world for a variety of reasons,however one particular constraint has been cost. The cost barrier is slowly being lowered which isallowing small and medium e-tailers (SMETs) and even individuals to more effectively compete inthe online market. Social networking is a prime example of these cost-reducing tools that aregrowing in usage. Social networking tools are now allowing SMETs to offer touch and feel productsthat are perishable over the internet. This research showcases this development in e-tailing of bakedgoods made possible through social networking and marketing strategies.

This paper examines the dynamics in the convergence of social networking, e-tailing andusers. The dynamics and strategies that are identified in this research are vital for SMETs toevaluate their online businesses, utilize a value-adding tool and effectively allow them to operatemore competitively in the global environment.

There is insufficient literature on customer to customer value creation online; so this studyunderscores the importance of C2C value creation in the dynamics of the convergence.

This paper is organized into 4 parts. We begin by describing Social Networks and SNS, E-tailing, and the members of the social network (individuals or companies). In part two, we discussrelevant theories and develop a study framework to analyze the dynamics of the convergence. Wethen look closely at two SMETs in the e-tailing business and evaluate their cases using theframework. Finally, we summarize the critical success factors and suggest strategies for e-tailingresulting from the convergence.

SOCIAL NETWORKING AND E-TAILING DEFINITIONS AND OVERVIEW

This section will describe social networks, internet retailer or e-tailer, and SNS members.Definitions and overviews will be provided for the concepts as well as an identification ofindividuals/companies that make up the memberships of social networks.

Social Networks

A social network is a virtual community where members with a common purpose or interestinteract online to share information, support and concern (Siegel, 2004). Social networks can alsobe considered as a web of personal relationships (Macaulay, Keeling, McGoldrick, Dafoulas,Kalaitzakis & Keeling, Summer 2009). They provide a venue for socialization and business wherefriends, co-workers, and business contacts create a virtual community where they can interact(Hopkins, 2003).

Social networks provide linkage between community members by allowing them to sharevirtually an unlimited, unrestricted amount of information through different forms and mediums.Common social networking forms include micro blogging, subscriptions, status, mobile text alerts,

67


blogs, instant messaging, and forums among others. Users primarily share these forms through themediums of the web and cell phones.

There are numerous social networking sites (SNS), but the top three, based on number ofvisits are Facebook.com (1,191,373,339), Myspace.com (810,153,536) and Twitter.com(54,218,731) (Kazeniac, 2009). As is evident from the visit counts, the growth of SNS has beentremendous. In the case of Twitter, users have jumped to an estimated 32.1 million from 1.6 milliona year ago (Vascellaro, 2009). These sites are setting the standard for the growth and the directionssocial networking is taking and can be used as good examples of their varied functionalities. Thefollowing are overviews of the top three SNS.

Facebook

Facebook was founded in 2004 and its mission is to give people the power to share and makethe world a more open and connected place. It also has over 400 million active users (based on usersrevisiting the site in the last 30 days) (Facebook). Facebook’s primary networking features that arerelevant and helpful to SMETs are:

! Facebook Connect: Provides access for Facebook on the web, I-phone and mobileweb. It has many features to help drive traffic, increase engagement, and growrevenue on the user’s website or application.

! Facebook Pages: Allows for profiles of brands, which look and behave like userprofiles, to connect and engage with customers and amplify messages to friends.

! Facebook Advertising: Gives users a do-it-yourself advertising service to reachtargeted audiences and connect real customers to SMETs. It can also assist in findingcustomers before they search.

MySpace

MySpace was launched in 2004 and brings together more than 100 million active usersaround the globe. Its focus is connecting people through personal expression, content, and culture.The firm empowers its global community to experience the Internet through a social lens byintegrating personal profiles, photos, messaging, games, and the world's largest music community(My Space Fact Sheet). Relevant and helpful MySpace’s networking features are:

! Basic Functionality: Includes friend search, customized personal profiling,picture/video upload, music search, creation of own web address, subscriptions,status, mobile text alerts, blogs, Myspace IM, and forums.

68


! MySpace My Ads: Provides users the ability to conduct banner advertising thatenables targeting, tracking and performance monitoring.

! Connect with MySpace: Increases social networking by allowing connectionbetween MySpace accounts and sites around the Web, desktop clients, mobileapplications, and more.

Twitter

Twitter began in 2006 and provides a real-time information network for people around theworld. Twitter is a pioneer in micro blogging where users keep others informed of their currentstatus by way of text messaging, instant messaging, e-mail, or the Web (My Space Fact Sheet).Users have unlimited use as long as their message is within 140 characters or less in limit(Chapman). It can be accessed through more than 50,000 third-party Internet and mobileapplications (Twitter for businesses). Twitter’s SMET applicable networking features are:

! Basic Tool: Allows a real-time, micro blogging connection with an audience toshare information.

! Customer Direct Interaction: Although part of the standard functionality, morecompanies are using Twitter to collect feedback on products or services,communicate advertising offers, build relationships and gather real-time marketintelligence (Twitter for businesses).

More and more, retailers are using social networking sites as channels to market orcommunicate information about their latest products (Moroz, 2008). SNS and online word of mouthhas even been identified as a threat to traditional marketing practices. Companies like Nike andHonda have built online communities where consumers have a public voice (Stern & Wakabayashi,2007), promotional product videos are seen in Youtube.com, and use of Twitter to inform customersabout online sales and discounts (Wade, 2009).

SNS poses both advantages and disadvantages for SMETs using the services for marketingor communication. Advantages include helping e-tailers keep track of what is said about theirbrands and delivers business value when customers need it most (King, 2008). It also assists withtransparency and in creating a cohesive corporate identity (Chen & Leteney, 2000; King, 2008) aswell as helping to build brand awareness (Chapman). A primary function of SNS allows the SMETsto build on-line “brand” communities which can add stability and continuity to businesses(Macaulay, Keeling, McGoldrick, Dafoulas, Kalaitzakis & Keeling, Summer 2009) while makingconnections, identifying prospective customers and pointing others to a SMET’s website or otherwebsites for resources. Lastly, the SNS provided search tools can significantly increase a SMETsexposure within their targeted industry or area of interest.

69


As a disadvantage, SNS can also provide a platform to e-tailers who are impersonating aSMET (domain squatting) to send out messages that were not authorized by the company (King,2008). This impersonation can confuse a SMET’s marketing message and can lead to loss ofbusiness and downgrades in consumer trust.

Internet Retailer or E-Tailer

Internet retailing or e-tailing is the newest store format in retailing (Grewal, Iyer & Levy,2004). An e-tailer can be classified as a pure play e-tailer or a bricks and clicks e-tailer. A brickand click e-tailer has a physical store and uses the internet whereas a pure play e-tailer only uses theInternet to sell their goods. This study focuses on e-tailers that use pure play as their initial andprimary approach to the market.

Pure play e-tailers pose one of the most interesting possibilities for e-tailing’s future as theyare identified as being small and engaged only in Internet selling. These e-tailers cover variousniches and hard to find products and have a tendency to be created and run by SMEs (Grewal, Iyer& Levy, 2004). The internet market has been conducive to SME operated businesses as it can levelthe competitive playing field by allowing small companies to extend their geographical reach andsecure new customers in ways formerly restricted to much larger firms (Chong, 2008). Therecession has accelerated the trend of a fast growing e-tail sector of small businesses and individualswith online shops with the prediction that global online sales will grow by over 10% this year(Solomon, 2009). More e-tailers means more competitive offerings for consumers.

This growth can be attributed to a number of benefits. E-tailing benefits entail an ease ofcommunication between company and consumer (Chen & Leteney, 2000), few infrastructurerequirements (Enders & Jelassi, 2000; Hofacker, 2008), a ubiquitous convenience, as well as agreater reach and accessibility or a potential worldwide reach thus allowing greater sales to SMEs(Enders & Jelassi, 2000; Grewal, Iyer & Levy, 2004). In addition, e-tailing provides consumers witheasy access to information on their assortment of goods, competitors’ product offerings, relatedinformation for better purchase decisions (Chen & Leteney, 2000; Grewal, Iyer & Levy, 2004) anda high level of privacy and anonymity in the purchase of certain sensitive products (Grewal, Iyer &Levy, 2004). Combining all of these benefits provides an extremely conducive utility for consumersand companies alike.

E-tailing is not without its issues. For an e-tailer, operational issues range from determininghow much of a product range or information to make available to consumers (Chen & Leteney,2000), high shipping and handling costs, problems with economies of scale due to small numbersof repeat buyers and small orders (Grewal, Iyer & Levy, 2004) and the need for an efficientdistribution system (rapid response demanded by Internet customers) (Chen & Leteney, 2000;Grewal, Iyer & Levy, 2004). Distribution is a key factor as it can avoid the risk of lost shipmentsand fraudulent claims (Chen & Leteney, 2000), but requires complex coordination which may result

70


in late deliveries (Enders & Jelassi, 2000). Other ssues include security problems (credit cardpayments), general lack of technical expertise, and weak institutional arrangements (Chong, 2008).

Primary consumer issues with e-tailing are the result of an absence of the physical interactionand shopping experience (Enders & Jelassi, 2000), the lack of a pre-trial experience, the lack ofinterpersonal trust for products that are difficult to evaluate in terms of value, as well as a lack ofinstant gratification (Grewal, Iyer & Levy, 2004).

One of the newest additions to the application of e-tailing is its linkage with SNS or viceversa. To address the issues brought forward, e-tailers tie up with social networks (or vice versa)to create a platform that provides the ability for consumers to receive advice and recommendationsfrom their social network friends. SNS provides this capability for e-tailers to enhance and do betterbusiness and effectively transform an internet shopping experience from a solo to a social activity(MacMillan, 2009). The popularity of SNS has led to further integration with the Internet industry(Rubel, 2009). According to Farquhar and Rowley (2006) the organizations that control the mostpopular virtual communities can dominate on-line business transactions.

SNS Members (End Users/Individuals/Companies)

Members of a social network provide and share personal information and views and expectlegitimate wide networking features that would cover music, video, groups, etc. Members also wantto network in a secure and easy environment (Social Networking Websites Review). SNS membersmay be individuals, groups, brands or business entities. In general, members of SNS sites bringtogether resources that create powerful capabilities which, in turn, create new sources of value forother member users (Hanson, 2000).

SNS provides its users many advantages and disadvantages. For end-users or individuals,SNS provide them with a public voice (Stern & Wakabayashi, 2007), freedom of expression (Keller& Libai) and a growing amount of consumer sovereignty (Grewal, Iyer & Levy, 2004). It can alsoprovide them the ability to interact with homepage content and with other users, empower them tobe producers of communcation (Asberg), and generate user content (Wade, 2009).

In terms of their shopping experiences, SNS can assist by provding more information onproducts as well as support increased social purchase involvement and a more intense shoppingexperience (Macaulay, Keeling, McGoldrick, Dafoulas, Kalaitzakis & Keeling, Summer 2009). Inaddition, SNS provide end users with increased consumer welfare through improved economicefficiency and greater choices (Grewal, Iyer & Levy, 2004). These same advantages can also leadto disadvantages. End users are subject to personal indecision due to the myriad of choices (Zmuda,2009), lack of instant gratification, lower customer service based on the lack of high-contact pre-and post sales service, increased difficulty in product returns, exchanges and monetary refunds, aswell as potential security issues (encryption, network and transactional privacy and security)(Grewal, Iyer & Levy, 2004).

71


E-tailing companies also receive many benefits from SNS. For example, it can help facilitatemarketing, communication and interaction with customers as well as providing options for brandingand even engage customers by providing a real-time, powerful and efficient two-way dialogue orcomplaint management between customers and the business (King, 2008; Martin, 2009; Pande,2008). SNS also provides a mechanism to reach a wider sphere of demographics since users areusually grouped according to their interests (Garrahan, 2009). It can also be a useful channel formarket research which targets different groups such as young consumers and other trendsetters thatcan influence the success/failure of a brand (Asberg).

STUDY FRAMEWORK

Theoretical Framework

Since its inception, e-tailing has been the subject of research interest regarding an entirespectrum of subjects from fulfillment efficiency to new marketing approaches. With the statedobjective of this study being the examination of the dynamics in the convergence of socialnetworking, e-tailing and users, three theories are used to develop the conceptual model.

Updated communications model

The first theory is the updated communications model. This theory helps to illustrate howtechnological and social developments are shaping communications. Senders are not in completecontrol of the messages that they send. Receivers are now playing more active roles incommunication by helping shape the messages they receive. This development shows the increasinginability of the sender to control the media environment.

Figure 1 Updated Communication model

Sender

Receiver

Sender Communication

medium

Receiver

Sender

receiver

(Solomon, 2009)

72


The new media environment has led to new message formats pertaining to e-tailing, m-commerce and blogging. M-commerce (mobile commerce) is a message format where marketerspromote their goods via wireless devices (i.e. cell phones). Blogging is when people post messagesin two basic forms to the Web in diary form. The first blogging form is RSS (real simplesyndication) where people sign up to have updates sent automatically to their computers. Thesecond form is twittering where people share moment by moment reports on what they are doingthrough Twitter.com (Solomon, 2009). This updated communication model provides the basicfoundation for SMETs to utilize social networking to achieve improved customer communication.

Value of e-commerce to the customer

The second theory relates to the value of e-commerce to the customer (Keeney, 1999). It hasbeen identified that e-commerce offers customers the net value of the benefits and cost (valueproposition) of both a product and the process of finding, ordering and receiving it. Keeneyidentified eleven points of value that internet commerce provides to consumers. These points are:

1. Overall objective: Maximize customer satisfaction2. Maximize product quality3. Minimize cost: Product, tax, shipping, internet,

and travel4. Minimize time to receive product: Delivery,

shipping, and dispatch5. Maximize convenience: Purchasing, time, quality

after sales service, easy return process, lessshopping effort, and ease of finding product

6. Minimize time spent: Purchasing, processing,payment, queuing, f inding products,communicating, searching, ordering and selecting.

7. Maximize privacy: Avoid electronic mailinglists

8. Maximize shopping enjoyment9. Make shopping a social event – minimize worry,

inspire customers, enhance user productivity,minimize regret, minimize disappointment,maximize customer confidence, reduce demandfor forced labor

10. Maximize safety: maximize driving safety,minimize risk of product use

11. Minimize environmental impact: Reduceenvironmental damages and pollution

Through these eleven points, customers will choose offerings that provide them the highestlevel of value, achieved through the combination of the product’s benefits and price and based ontheir personal beliefs and preferences.

Word of mouth (WOM)

The last theory is word of mouth. Word of mouth (WOM) can be defined as informal oralcommunication between individuals, companies, etc. (Silverman, 2001). WOM communication hasseveral sources and properties and possess a great deal of intrinsic power. Silverman (2001)identified a number of functionalities and characteristics of WOM which are shown in Table 1.

73


Table 1: Sources of Word of Mouth

Sources of word of mouth Function Content (what it provides)

Company advocates Information Claims, benefits

Experts Confirmation Upside/downside potential under the best circumstances

peers Verification What to expect in the real world, in typical situations

(Silverman, 2001)

WOM communication is customer driven, self-generating through exponential growth,unlimited in speed and scope and possesses a fair amount efficient and time-saving communicationvalue.

Online WOM creates buzz. Buzz is described as the result of the online interaction amongconsumers that serves to amplify the original marketing message (Thomas Jr., 2004). It ischaracterized by being covert, grass roots, authentic, credible (Solomon, 2009) and provides e-tailerswith a sense of:

! Ubiquity with a communications network accessible by all! Expectations by showing the importance of network participants’ view of successful

technologies! Sharing the efficiency that comes when a network allows reciprocity! Specialization from the network allowing firms and providers the ability to specialize

in their own key capabilities ! Virtual value activities that the network delivers through useful information (Hanson,

2000).

These three theories form the foundation of the conceptual framework that will be used forthe analysis of the two case studies that demonstrate the convergence of social networks, e-tailersand users.

Framework for Analysis

The Internet is a nearly perfect market because information is instantaneous (Kuttner, 1998).E-tailers have a venue to market, communicate, interact and deliver service to their customersthrough their websites or SNS link and consumers can compare offerings of sellers worldwide. SNSprovides the facility for all users , consumers and e-tailers, to interact online, share information andconcern. The convergence of the e-tailer, consumer, and SNS in the web is providing synchrony andencouraging online interactivity.

74


This convergence creates online WOM which creates the buzz. The properties of the buzzmake online WOM relevant for SMETs: it can address a consumer’s lack of experience with aproduct and lower the perceived risk towards buying the product; it functions as the source ofinformation from e-tailers; it facilitates confirmation and verification from peers/experts regardingproducts, costs, and the convenience in finding, ordering and receiving products, risk in privacy andsafety; and it enables the consumer to have a social shopping experience.

Figure 2 Convergence of SNS, E-tailers and Users

This convergence results in the framework used by this study to analyze the presented businesscases.

Two cases will be analyzed using the convergence framework. The elements in theframework that will be evaluated are the website, the SNS link, member users and the buzz.

CASE STUDIES: SMET UTILIZATION OF SOCIAL NETWORKS

This research utilized two case studies in its identification of the dynamics of theconvergence of social networking, e-tailing and users. The first study concerns an e-tailer of creampuffs and other baked goods, Mommy Puff Bakeshop. In a similar vein, Hot Blondies, the secondcase e-tailer, is also a producer and marketer of baked goods. Both firms operate e-tailing websitesand utilize social networking as part of their sales, marketing and communications strategies.

Mommy Puff Bakeshop

Mommy Puff Bake Shop was launched in 2003 and currently makes and sells pastries likecream puffs, éclairs, and caramel walnut tarts. Since its founding, the firm has grown and openeda retail outlet in addition to its online sales activities. Despite its growth, Mommy Puff Bake Shopremains a basic example of a SMETs first presence on the internet.

75


Mommy Puff’s website can be found at http://mommypuffbakeshop.webs.com and providesbrochure level information about the firm. The web site lists information showing rates and service,About Us, guestbook and contact information. Rates and service shows pictures of products,indicates that they are made to order (e-mail) and pickup/delivery areas. This section also gives their6-item menu, minimum order quantities and prices. Contact information is also provided givingusers access to the brick-and-mortar address, landline, cell, website and Facebook link. About Usinformation consists of a brief history of both the on- and off-line shops, incomplete contactinformation of a partial address, landline, cell and email address. The guest book allows visitors tomake comments about their purchases and for the site owner to discuss product information. Thecomments span a period from 14 February 2009 to 5 October 2009. Finally, the contact pageprovides the same information as the About Us page, but does allow a user to send an integratedemail to the bakery owners.

Mommy Puff utilizes two virtual communities to execute its SNS strategy. The first socialnetworking site used is Multiply.com which acts as an extension of their web page and providesbasic information about the bake shop that includes contact information (complete address, landlineand cell numbers), shop background and the opening of a retail outlet as well as a product listcomplete with photos, packaging, quantity and pickup price. In addition, the site also provides 3 RSSfeeds with the latest dating 13 March 2010 approximately a month before the study’s informationwas gathered. The RSS information outlined contact information, a brief background of the shop anda product list consisting of 11 items indicating that the products are for pickup only.

The second community utilized by Mommy Puff is Facebook. Facebook contains much ofthe same basic information as Multiply in terms of company overview and background. It differsslightly in that administrators and links (Twitpic, Multiply and their website) are listed and morephotos are displayed. In addition, their site lists 427 members, an old discussion, and a wall with22 posts from February 2009 to April 2010.

Hot Blondies Bakery

Hot Blondies is a bakery that sells a wide range of brownies and a t-shirt via the Web,selected retailers, and events, in addition to catering. This example case provides a professionallycreated and well designed website coupled with a well organized and diversified business plan thatis a large step past that of the first introduced case.

Hot Blondies website can be found at http://www.hotblondiesbakery.com and also providesbrochure level information about the firm in addition to direct e-shopping functionality. Thebranded website provides the basic abilities to view their products, the shop, and read about thefirm’s news, catering options, FAQs and a brief history. The shopping section allows users fullaccess to photos and descriptions of their 14 types of baked goods and a t-shirt with multiplepayment types and delivery options. Although contact information is limited to a phone number and

76


email address, provided news items notify users of where the products can be purchased outside ofthe Internet at regular functions and physical stores. These new items span a period from 22 January2009 to 4 May 2010 and allow their followers to be up to date on recent news and events.

Hot Blondies uses only Facebook in its SNS strategy. Their Facebook site contains similargeneral information as can be found in their website, such as branding, photos, contact data and anoverview, but also allows for more customer interaction. The general information covers contactinformation such as email address, city and URL as well as displaying a much greater number ofproduct shots than from their website and a dozen events from 2009. The e-tailer’s Facebook pagealso allowed for more interaction in showing over 550 members, discussion and wall posts from 30June 2008 to 21 April 2010.

CASE STUDY EVALUATION USING THE FRAMEWORK

Mommy Puff Bakeshop

Evaluating Mommy Puff through the identified framework requires an examination of thebakeshop’s three different web pages and social networking sites in terms of looking at theconvergence between their virtual tools’ provided information and the buzz factor generated by theconfirmation/verification qualities.

First, examining Mommy Puff’s information provided across its web and SNS platforms, itis evident that they have not made a concerted effort to regularly update and synchronize thecontents. In particular, there are a great deal of discrepancies regarding delivery options, links,addresses and contact information. An additional neglected area of information is that regardingsecurity and privacy, however, as their website and SNS sites do not facilitate ordering, it isunderstandable. A correlation of information about Mommy Puff can be found in Table 2.

Table 2: Mommy Puff Bakeshop: Correlation of Information from the e-tailer

Content Website Inform Multiply Inform Facebook Inform

Product/Cost 6 items 11 items 11 items

Made to order

Product Photos 13 15 16

Convenience to find Complete address Complete address withinstructions

Incomplete address

Convenience to order By email or leave phonenumber

Call By email or leave phonenumber

Convenience to order Email indicated Email Not indicated Email indicated

Convenience to order 2 landline numbers 2 landline numbers 2 landline numbers

77

Table 2: Mommy Puff Bakeshop: Correlation of Information from the e-tailer

Content Website Inform Multiply Inform Facebook Inform


Convenience to order 2 cell phone numbersdifferent from SNS

1 cell phone number 1 cell phone number

Convenience to receive Delivery in 3 areas or pick up Pick up Pick up

Risk in privacy/ safety

Social experience Indicates Facebook but nolink

RSS feed Link to website, Twitpicand Multiply

Social experience 16 posts from guestbook 3 RSS feeds 22 wall posts

The two-way interaction between the e-tailer and customers allows confirmation andverification to occur. The guest book facility of the website, RSS feeds of Multiply.com and the wallposts of Facebook enable interaction between Mommy Puff and her customers as well as initiatingconfirmation and verification between peer and expert customers. The confirmation and verificationof product information provides for the “buzz” factor in the convergence between the SNS, user andthe e-tailer. Table 3 shows the buzz among the three groups.

Table 3: Mommy Puff Bakeshop: Buzz Evaluation of Posts from E-tailer, Peers and Experts

content Buzz Function (Website) Buzz Function (Multiply) Buzz Function (Facebook)

Customer Value Inform Confirm/ Verify Inform Confirm/ Verify Inform Confirm/ Verify

Product 6 1 1 8

Cost 1 1 3

Convenience in finding 1 1 1 3

Convenience in ordering 1 2 1

Convenience in receiving 2 3

Risk in privacy and safety

Social experience. 1 4 5 1

Most recent post Oct. 15, 2009 Mar. 14, 2010 April 2, 2010

Hot Blondies Bakery

Hot Blondies’ evaluation also requires analyzing their web page and social networking sitefor provided information and buzz factors. In looking at their web site, it is evident, given theprovided credit information, that they have had professional help building and maintaining their web

78


presence. In addition, although their Facebook page displays their web link, Hot Blondies’ webpage shows no indication of their presence on an SNS. Furthermore, despite their active updates,there is a need for synchronization between their Facebook and web pages. A correlation ofinformation about Hot Blondies can be found in Table 4.

Table 4: Hot Blondies Bakery: Correlation of Information from the E-tailer

Content Information in Web Site Information in Facebook

Product/Cost 15 items Indicated products

Convenience to find Indicated city and landline Indicated city no landline

Convenience to order Shopping cart Through email

Convenience to order Indicated email Indicated email

Convenience to receive Shipping and delivery page

Risk in privacy and safety Specifies site maintenance and consultants

Social experience Link to facebook not indicated Website link indicated

# of posts: 34 wall posts; 2 discussion posts

Most recent post April 20, 2010 April 20, 2010

As there was no guest book on the Hot Blondies website, all buzz is generated throughFacebook wall posts. The wall posts enable customers and members to interact with the e-tailer aswell as confirm and verify information with other customers. Table 5 shows the buzz between thise-tailer’s 2 groups.

Table 5: Hot Blondies Bakeshop: Buzz Evaluation of Posts from E-tailer, Peers and Experts

Content Buzz Function (Website) Buzz Function (Facebook)

Customer Value Inform Confirm/Verify Inform Confirm/ Verify

Product 6 9

Cost 2

Convenience in finding 3

Convenience in ordering

Convenience in receiving 2

Risk in privacy and safety

Social experience.

Most recent post April 20, 2010

79


Buzz Analysis of Case Information, Confirmation and Verification

As introduced in the framework, buzz functions as a source of information from e-tailers,as well as the confirmation and verification from peers/experts regarding: product, customer cost,convenience, risk and social experience.

The first area regards the actual products offered in this case. Especially in the past, therewere limitations on the types of products that could be offered online. Digitized materials, such assoftware or music, were considered more conducive to internet sales (Enders & Jelassi, 2000).However, in these cases, despite the high perishability of the products, both SMETs were notrestricted from using the Web as their venue. They instead utilized the capabilities of the Web anddeveloped appropriate strategies. In the case of Mommy Puff Bakeshop they designated deliverypoints, delivery schedules, minimum orders, and pre-ordering. In the case of Hot Blondies, they alsoidentified delivery schedules, minimum orders and pre-ordering.

Customer cost is another area involving buzz. SMETs are faced with three areas where costshave a great effect. The first is customer cost bound by the lack of trial for products that require highsensory evaluation (Grewal, Iyer & Levy, 2004). Although it is impossible to give customers a biteor smell from their products, the bakeshops used technology to bridge that gap. Mommy PuffBakeshop utilized both website and SNS to confirm and validate their products and the productexperiences of peers/customers. On the other hand, Hot Blondies Bakeshop only utilized their SNS.

Shipping and handling costs are another factor facing e-tailers (Grewal, Iyer & Levy, 2004).These costs are not an issue for Mommy Puff, since they have schedules in delivering in designatedareas, but may be an issue for Hot Blondies. The last customer cost issue is the result of the higheconomies-of-scale due to low repeat buys and small order sizes (Grewal, Iyer & Levy, 2004). Inthe case of Mommy Puff, the identified delivery and pick up points only partially covers this issue.In the case of Hot Blondies, their shopping cart facility and zone pricing more thoroughly coversthis aspect.

Convenience is the third variable that would be covered by the buzz. Consumers typicallyenjoy a high degree of instant gratification when shopping in traditional stores. The required waitwhen shopping via the internet is a limitation on receiving that gratification (Grewal, Iyer & Levy,2004). In both e-tailing cases, this need for instant gratification depends on the customer. If theywant to immediately receive the product, they can choose to pick it up from a delivery point orphysical store distributor rather than wait for product delivery.

Another area of convenience is that of customer service. Customer service from an e-tailerlacks the high-contact pre- and post-sales service (high agent to customer ratio) (Hanson, 2000).This poses limitation in terms of: product returns and exchanges and monetary refunds (Grewal,Iyer & Levy, 2004); types of interaction (real time, messaging); and type of service (self) (Hanson,2000). In the case of Mommy Puff, the creator seems to have a “favorite” in terms of where toupdate. Thus, some venues, like their website, are quite outdated. This channel management further

80


aggravates the issue of lack of high contact. However, establishing a retail outlet for some customerswould partially cover this channel management flaw. In the case of Hot Blondies, they regularlypost current information about their activities as an e-tailer so even if there is lack of personalcontact, the customer is assured that the website is regularly checked and can be relied upon as avalid medium. Hot Blondies has also identified selected retailers to carry their products, thus, morethoroughly covering this issue.

The next area to investigate is risk. These days, the need for encryption, network security,and transactional privacy and security as a means to ensure security and privacy have become aparamount concern (Grewal, Iyer & Levy, 2004). In this regard, network security would not be theconcern of Mommy Puff customers since no financial transactions can happen on their web site.However, there may be an issue with regards to customers providing their names, addresses andtelephone numbers when they order. In the case of Hot Blondies, although they have sitemaintenance and a consultant to provide them the security needed for online transactions, justlooking at their website does not visually assure consumers of solid privacy and security.

The final area regards the social experience customers receive from their online shoppingexperiences. In both cases, personal indecision in terms of product, cost, convenience in receiving,risk in privacy and safety is confirmed and verified through SNS.

SMET SOCIAL NETWORK CRITICAL SUCCESS FACTORS

As demonstrated through the cases, a number of critical success factors have been identifiedthat allow a virtual community, when incorporated into a business strategy, to enhance the businessof an e-tailer. The seven critical success factors include: access to relevant synchronizedinformation, personalization of communications, trust-building, lengthened reach, additionalmarketing or communication channel, additional feedback channel, and a multi-purpose method formotivation fulfillment.

The first critical success factor regarding improved customer access to information stemsfrom the incorporation of a virtual community into an e-tailer’s business strategy. To be successful,e-tailers need to provide accessible, relevant and synchronized information on product, cost, orderand delivery terms, privacy and safety in doing Internet transactions. SNS is critical for bothproviding a full range of information and assisting in providing synchronized and up-to-dateinformation. Synchronicity and relevance is critical, especially when the primary medium ofcustomer interaction is virtual. Between the two cases Mommy Puff Bakeshop needs to be morevigilant since they have to synchronize information on 3 channels as compared to Hot Blondies whohas 2 channels and has professional help for website maintenance.

A second success factor comes from personalization of communications. It is necessary foran e-tailer to provide its customers a personalized transaction process. Immediate or real-timefeedback in websites and SNS would give the assurance that consumers are being attended to and

81


given importance. Converging channels (website and SNS) facilitate instantaneous communicationso delayed feedback from the e-tailer is frustrating to the user and at the same time visible to otherusers. Thus, these channels should be efficiently utilized by e-tailers. Mommy Puff again needsimprovement in this aspect as having a “favorite” channel compromises the other channel and postsin their website and SNS does gives the impression that there is nobody manning the store. In thecase of Hot Blondies, it helps that their posts are current and covers their marketing activities. Thisgives the impression that they are “actively marketing” and that their channels are manned.

Trust building is the third primary factor required for e-tailing success. Virtual communitiesencourage user participation and even foster genuine relationships with customers, thus creatingemotional links that boost brand differentiation and encourage product loyalty (Chong, 2008).Active posting of Mommy Puff, personal postings, even if it is not marketing related developsrelationship between e-tailer and customer. It is like opening up and reaching out to its customers.On the other hand posting of events by Hot Blondies showcases their marketing and social activitythat encourages member participation.

Virtual communities can reach large numbers of consumers. This reach is possible given thefunctionalities of the convergence that creates the buzz. The communications network ischaracterized by being covert, grass roots, authentic, credible (Solomon, 2009). Social networkingsites provide the means for businesses to securely reach different segments since the users areusually grouped according to their interests. They are now becoming a more popular choice foradvertising campaigns since they can reach about fifty million users daily (Garrahan, 2009).Evidence of this is in the number of SNS members registered to both companies in the providedcases.

The fifth success factor is connected to the virtual communities’ provision of anothermarketing and communication channel. Both cases make use of virtual communities to complementtheir websites in communicating to their customers.

Social networks can also act as feedback channels since they give users a venue for user-generated content, collaboration, interactivity, and information sharing (Wade, 2009). As a feedbackchannel users have the opportunity to voice out their feelings, beliefs and experiences. Bycollaborating, prospective consumers can get helpful information from other consumers and assistin overcoming personal indecisions. Interactivity in social networks makes users more involved.Information-sharing verifies and confirms what has been provided by e-tailers as to information intheir websites or SNS. Both cases utilize this functionality of their SNS which enable them to“observe” and interact with their customers.

Both satisfied and dissatisfied consumers can broadcast their views immediately which couldcreate “buzz” that could either be positive or negative (i.e. online activism) (D'Angelo Fisher, 2009).Both cases studied in this research utilize SNS as a feedback channel in addition to their telephones,cell phones and websites. In both cases, registered feedback is positive.

82


The final critical success factor is based on a virtual community’s ability to act as a multi-purpose method for motivation fulfillment. The integration of the internet industry exposes socialnetworkers/customers to varied situations that enable them to fulfill several or even multiplemotivations for visiting sites: product information and purchase; social support and information;while others are looking for a more intense experience and greater social involvement (Macaulay,Keeling, McGoldrick, Dafoulas, Kalaitzakis & Keeling, Summer 2009). An example where SMETscan benefit from this factor is when a manufacturer’s representative or service provider uses an SNSto positively interact with a customer, thus changing the customer’s perspective for the better (King,2008). This critical success factor is, again, true in both referenced cases.

Overall, many critical success factors that involve SNS have been identified. Properlyincorporating these factors into business strategies are critical to the survival and success ofmembers of the e-tailing community.

SMET SOCIAL NETWORK STRATEGIES

Based on the learnings from the explored cases, eight strategies can be identified in whichSMETs can make many value-adding uses of social networks and virtual communities. Thesestrategies use social networking to:

A. Promote relationships and build trust (Macaulay, Keeling, McGoldrick, Dafoulas,Kalaitzakis & Keeling, Summer 2009). SMETs can use SNS in conjunction with theirbrands/companies to create relationships through constant communication and/or interactionvia the online community.

B. Increase brand awareness and commitment (Macaulay, Keeling, McGoldrick, Dafoulas,Kalaitzakis & Keeling, Summer 2009).

C. Add value through the confirmation and validation from peers to other customers (Macaulay,Keeling, McGoldrick, Dafoulas, Kalaitzakis & Keeling, Summer 2009).

D. Integrate the e-tailer into a clicks and bricks company (Grewal, Iyer & Levy, 2004). As seenin both cases, e-tailers can open a physical front end with the customer base they created.

D. Pursue niche strategies. Both cases utilize this strategy in specifically indentifying targetcustomers who love cream puffs and brownies.

83


E. Better position their products and brands as current Internet technology does not effectivelyallow retailers to convey the status differentiation of their stores (Grewal, Iyer & Levy,2004).

CONCLUSION

Through the use of the two case studies, the dynamics in the convergence of socialnetworking, e-tailing and users have been more fully identified. In addition, six strategies wereidentified in the hope of aiding SMETs better evaluate their online businesses and provide them witha powerful, value-adding tool to allow them to operate more competitively in the globalenvironment.

REFERENCES

Acquisti, A. (2005). Information revelation and privacy in online social networks (The Facebook case). Proceedings ofthe ACM workshop on Privacy in the electronic society, 71-80.

Allen, S. (n.d.) Guide to online business networking. Retrieved February 6, 2009, from http://www.work.com/online-business-networking-767

Anon. Technology: we are all retailers (2008, May). Country Monitor, 1.

Applegate, L., C. Holsapple, R. Kalakota, F. Radermacher & A. Whinston (1996). Electronic commerce: buildingblocks of new business opportunity. Journal of Organizational Computing & Electronic Commerce, 6(1), 1-10.

Asberg, P. (n.d.). Using social media in brand research. Retrieved March 23, 2010, fromhttp://www.brandchannel.com/images/papers/433_Social_Media_Final.pdf

Awad, N. F. & A. Ragowsky (Spring 2008). Establishing trust in electronic commerce through online word of mouth:an examination across genders. Journal of Management Information Systems, 24(4), 101-121.

Cao, M., Q. Zhang & J. Seydel (2005). B2C e-commerce web site quality: an empirical examination. IndustrialManagement & Data Systems, 105(5), 645-661.

Capineri, C. & T.R. Leinbach (2003). Globalization, e-economy, and transport. Growth & Change, 34(4), 385-389.

Chapman, C. (n.d.). Brand Awareness: It’s all a ‘Twitter. Retrieved March 23, 2010, fromhttp://www.brandchannel.com/images/papers/493_Twitter_Brand_Ascension.pdf

Chen, S. & F.Leteney (2000). Get real! Managing the next stage of internet retail. European Management Journal,18(5), 519-528.

84


Cheung, C., M. Lee & N. Rabjohn (2008). The impact of electronic word-of-mouth: the adoption of online opinions inonline customer communities. Internet Research, 18(2), 229-247.

Chong, S. (2008). Success in electronic commerce implementation: a cross-country study of small and medium sizedenterprises. Journal of Enterprise Information Management, 21(5), 468-492.

D'Angelo Fisher, L. (2009, September). Bad word spreads. BRW, 47.

Enders, A., H. Hungenberg, H. Denker & S. Mauch (2008). The long tail of social networking: revenue models of socialnetworking sites. European Management Journal, 26(3), 199-211.

Enders, A. & T. Jelassi (2000). The converging business models of internet and bricks-and-mortar retailers. EuropeanManagement Journal, 18(5), 542-550.

F a c e b o o k ‘ D e t a i l e d I n f o ’ . ( n . d . ) R e t r i e v e d M a r c h 8 , 2 0 1 0 , f r o mhttp://www.facebook.com/facebook?ref=pf#!/facebook?v=info&ref=pf

Farquhar, J. & J. Rowley (2006). Relationships and online consumer communities. Business Process ManagementJournal, 12(2), 162–177.

Ferguson, R. (2008). Word of mouth and viral marketing: taking the temperature of the hottest trends in marketing.Journal of Consumer Marketing, 25(3), 179-182.

Garrahan, M. (n.d.). My Space on the money in social networking. Retrieved February 9, 2009, fromhttp://www.ft.com/cms/s/0/6317a9ec-ecc8-11dd-a534-0000779fd2ac.html

Grewal, D., G. Iyer & M. Levy (2004). Internet retailing: enablers, limiters and market consequences. Journal ofBusiness Research, 57(7), 703-713.

Hanson, W. (2000). Principles of Internet Marketing. Australia, Cincinnati, Ohio: South Western College Publication.

Hofacker, C. (2008). E-tail constraints and tradeoffs. Direct Marketing: An International Journal, 2(3), 129-143.

Hopkins, J. (2003, December). Investors court social-networking sites. USA Today, 3b.

Francis, J. (2007). Internet retailing quality: one size does not fit all. Managing Service Quality, 17(3), 341-355.

Kazeniac, A. (2009, February). Social Networks: Facebook takes over top spot twitter climbs. Retrieved November 30,2009, from http://blog.compete.com/2009/02/09/facebook-myspace-twitter-social-network

Keeney, R. (1999). The value of internet commerce to the customer. Management Science, 45(4), 533-542.

Keller, E. & B. Libai (n.d.). A holistic approach to the measurement of WOM its impact on consumer's decisions.Retrieved March 23, 2010, from http://www.slideshare.net/kellerfay/a-holistic-approach-to-the-measurement-of-wom

85


King, R. (2008, September) How companies use Twitter to bolster their brands. Retrieved November 27, 2009, fromhttp://www.businessweek.com/technology/content/sep2008/tc2008095_320491.htm

Kuttner, R. (1998, May) The net: a market too perfect for profits. BusinessWeek, 20.

Lacy, S. (2008, December). Time for E-Commerce 2.0. Retrieved November 27, 2009, fromhttp://www.businessweek.com/technology/content/dec2008/tc20081210_092729.htm

Lee, L. (1999, October). Clicks and mortar at Gap.com. BusinessWeek, 53-77.

Macaulay, L., K. Keeling, P. McGoldrick, G.Dafoulas, E. Kalaitzakis & D. Keeling (Summer 2009). Co-evolving e-tailand on-line communities. International Journal of Electronic Commerce, 11(4), 53–77.

MacMillan, D. (2009, October). Facebook banks on a little help from its friends. BusinessWeek , 48-49.

Martin, R. (2009). Web 3.0: leveraging social networking. Retrieved February 9, 2009, fromhttp://www.dealerscope.com/article/there-right-ways-wrong-ways-401970/

Mazursky, D. & Y. Schul (2000). In the aftermath of invalidation: shaping judgment rules on learning that previousinformation was invalid. Journal of Consumer Psychology, 9(4), 213-222.

McCormick, A. (2007, January). Social sites cash in on users via etail. New Media Age, 1.

Moroz, Y. (2008, April). Social networking evolves into venue to target shoppers. Retailing Today, 31-32.

My Space Fact Sheet. (n.d.) Retrieved March 8, 2010, from http://www.myspace.com/pressroom?url=/facts+sheet/

Pande, S. (2008). Social Networking: waiting to explode. Retrieved February 5, 2009, fromhttp://businesstoday.intoday.in/index.php?option=com_content&Itemid=1&task=view&id=8104&sectionid=25&issueid=41&page=archieve

Ratchford, B. T. (2001). The economics of consumer knowledge. Journal of Consumer Research, 27(4), 397-411.

Ratchford, B.T., D. Talukdar & M. Lee (2007). The impact of the internet on consumers’ use of information sources forautomobiles: a re-inquiry. Journal of Consumer Research, 34(1), 111-119.

Rubel, S. (2009, April). Who will rule the web once Twitter and Facebook fade? Advertising Age, 30.

Siegel, C. (2004) Internet Marketing Foundations and Applications. Boston: Houghton Mifflin Company..

Silverman, G. (2001). The secrets of word of mouth marketing: how to trigger exponential sales through runaway wordof mouth. New York: Amacom.

Smith, N. (2009, October). Setting out your stall. Consumer Etail, 18-19.

86


Social Networking Websites Review. (n.d.) Retrieved March 24, 2010, from http://social-networking-websites-review.toptenreviews.com/

Solomon, M. (2009) Consumer behavior buying having and being (Eigth Edition). New Jersey: Pearson Prentice Hall.

Stern, A. & C. Wakabayashi (2007). Are you ready for Web 2.0 Marketing? J@pan Inc, 72, 6-9.

Thomas Jr., G. M. (2004). Building the buzz in the hive mind. Journal of Consumer Behaviour, 4(1), 64-72.

Twitter for businesses. (n.d.) Retrieved March 8, 2010, from http://twitter.com/about

Vascellaro, J. (2009, May) Twitter trips on its rapid growth. The Wall Street Journal, B1.

Wade, J. (2009). The new WILD WEST. Risk Management, 56(8), 26-31.

Wiese, B., J. Sauer & B. Ruttinger (2004). Consumers' use of written product information. Ergonomics, 47(11),1180–1194.

Zmuda, N. (2009, April). Online shopping goes from solo experience to social interaction. Advertising Age, 15.

87


USE OF CREDIT CHECKS IN EMPLOYEE SELECTION:LEGAL AND POLICY ISSUES FOR EMPLOYERS

Gerald E. Calvasina, Southern Utah UniversityRichard V. Calvasina, University of West Florida

Eugene J. Calvasina, Southern University

ABSTRACT

A large number of U.S. employers use credit checks in making employee selection decisions.A recent Society for Human Resource Management (SHRM) study reported that 47 percent ofrespondents only use credit background checks for job candidates in certain types of positions, forexample, job candidates for positions with fiduciary and financial responsibilities (SHRM, 2010).Thirteen percent of those surveyed conduct credit checks on all candidates while 40 percent of therespondents in the SHRM study reported that they did not utilize credit checks on any job candidates(SHRM, 2010). In recent years, a broad coalition of opposition to their use has surfaced. Thecoalition includes advocacy groups like the Lawyers’ Committee for Civil Rights Under the Law,the U.S. Equal Employment Opportunity Commission (EEOC), state legislatures, and the U.S.Congress (Maurer, 2010). The purpose of this paper is to examine the use of credit checks byemployers in making employee selection decisions, the recent efforts to regulate employer use ofcredit histories in employee selection decisions, and the policy and practice issues for employers.

INTRODUCTION

Employers utilize a variety of tools to examine job applicants and candidates for promotionin making their decisions. Employers will typically contact references, verify educational orprofessional history, obtain a report on an individual’s criminal history, and in certain situations,obtain a report on an individual’s credit history (SHRM, 2010). The use of credit checks has comeunder increased scrutiny in recent years by a variety of influential individuals and organizationsquestioning the impact of their use in regard to their discriminatory impact and their validity andreliability in general.

As to the use of credit checks by employers, a recent Society for Human ResourceManagement (SHRM) study reported that 47 percent of respondents only use credit backgroundchecks for job candidates in certain types of positions, for example, job candidates for positions withfiduciary and financial responsibilities (SHRM, 2010). Thirteen percent conduct credit checks onall candidates while 40 percent of the respondents in the SHRM study reported that they did not

88


utilize credit checks on any job candidates (SHRM, 2010). The SHRM study also found thatorganizations conducted credit background checks on job candidates for a variety of positions (seeTable 1). 91 percent of the companies surveyed checked the credit of job candidates for positionswith fiduciary and financial responsibility (e.g., handling cash, banking, accounting, compliance,technology) (SHRM, 2010).

Table 1: Categories of Job Candidates That Organizations Conduct Credit Background Checkson Categories % of Organizations Conducting Checks

Positions with Fiduciary & Financial Responsibility 91%

Candidates for Senior Executive Positions 46%

Candidates with access to highly confidential information 34%

Candidates who will have access to company or other property or in a position of financial trust 30%

Positions for which state law requires a background check 11%

Candidates who will have security responsibilities 9%

Positions involving national defense & homeland security 8%

Safety sensitive positions 5%

Candidates who will work with children, the elderly, the disabled & other vulnerable populations 3%

Candidates who will work in health care or with access to drugs 3%

Other 4%

Source: SHRM (2010). Background Checking: Conducting Credit Background Checks

The SHRM study also found that 65% of the organizations in their study reported they gavecandidates the opportunity to explain the results before a decision to hire or fire is made ifinformation discovered in the credit check might have an adverse effect on an employment decision(SHRM, 2010). Another 22% of the companies surveyed would allow the candidate to explain afterthe decision to hire or not hire had been made. Only 13% of respondents reported that they wouldnot at any time offer the candidate the opportunity to explain the results (HRM, 2010). Therespondents in the SHRM study reported a variety of negative information that would adverselyaffect their decision to extend a job offer (see Table 2).

Table 2: Negative Information That Would Most Likely Affect Decision to Not Extend a Job Offer

Current outstanding judgment(s) (e.g., lawsuit filed in court) 64%

Accounts in debt collection 49%

Bankruptcy 25%

High debt-to-income ratio 18%

89

Table 2: Negative Information That Would Most Likely Affect Decision to Not Extend a Job Offer


Foreclosure 11%

Tax liens 10%

Education-related debt 2%

Medical debt 1%

Other 3%

Source: SHRM (2010). Background Checking: Conducting Credit Background Checks

The most frequently cited negative information in Table 2 was the presence of a currentoutstanding judgment(s) (e.g., lawsuit filed in court) that was reported by 64% of respondents(SHRM, 2010). Table 2 presents all the types of information reported by respondents with thepercentage of respondents that would use that type of information.

There are four primary reasons reported organizations in the SHRM study utilizing creditchecks. They include: (1.) to reduce/prevent theft and embezzlement, or other criminal activity by(54%) of respondents; (2.) to reduce legal liability for negligent hiring (27%) of respondents; (3.)to assess the overall trustworthiness of the job candidate by (12%)of respondents; and (4.) tocomply with applicable state law requiring a check for a particular position by (7%) (SHRM, 2010).These reported reasons are widely cited in practitioner literature and the popular press. Thefollowing two quotes are often observed:

Table 3: Often Observed Quotes

“When you think about it, people who have good credit keep their promises and are responsible, so it made sense thatif the credit was good, the honesty would be better”(Babcock, 2003).

“The consideration of credit history information can not only be useful in determining whether the potential hire hasthe skills and responsibility necessary for a particular job, but also whether the individual is qualified to handle money”(Bates, 2009).

Another underlying reason for the increased use of background and credit checks are reportsin the literature on the increased propensity of job applicants to provide inaccurate and often timesfalse information in their resumes and job application forms. Accu-Screen Inc. reported “spikes”in “resume falsification data” that they concluded are associated with “economic downturns andweak labor markets” (Minton-Eversole, 2008). In other studies cited by Minton-Eversole, reportsof up to 73 percent of individuals conducting background checks identifying lies in informationsubmitted by job applicants (Minton-Eversole, 2008). The propensity for job applicants to submitinaccurate and false information when applying for jobs is not new. In a 2003 HRMagazine article,the author cited studies that reported finding 44 percent of applicants lied about their work histories,41 percent about their education, and 23 percent falsified credentials or licenses (Babcock, 2003).

90


The suspect reliability of information being supplied by job applicants to employers and theincreased concern associated with negligent hiring are strong rationale to supporting the search formore accurate information on which to base hiring and promoting decisions.

Barbara R. Arnwine, Executive Director Lawyers’ Committee for Civil Rights Under Law,testified at a 2006 meeting of the Equal Employment Opportunity Commission (EEOC) meetingon race and color discrimination also noted the wide spread use of credit checks by employers inboth the public and private sector (Arnwine, 2006). Adam T. Klein’s statement at a May 2007meeting of the (EEOC) reiterated Arnwine’s observations and cited 1996 and 2003 SHRM studiesthat demonstrate that the use of credit checks and the practice of investigating backgrounds ofpotential employees has been on the rise in recent years. Klein went on to note that “a substantialand, in recent years, increasing number of employers are screening out job applicants for having anegative credit history” (Klein, 2007).

CURRENT FEDERAL REGULATION OF CREDIT CHECKS

The primary regulation of employer use of credit checks in the employee selection processcomes from the Fair Credit Reporting Act (FCRA). FCRA (15 U.S.C. § 1681 et seq.) was enactedin 1996 and most recently amended by the Title X of the Wall Street Reform and ConsumerProtection Act, Public Law 111-203. Employment provisions associated with employer use of creditreports in the selection process are posted at the Federal Trade Commission’s (FTC) web sitehttp://www.ftc.gov/ . One key provision for employers begins with the requirement for employersto get permission from an individual before requesting a copy of that individual’s credit report froma credit reporting company. This cannot be accomplished by simply requiring an applicant to signhis or her application for employment. The notice and authorization must be on a stand-alonedocument. If an employer utilizes the information from the credit report to take an “adverse action”such as a denial of employment or a promotion, a decision to terminate the employee or a decisionto reassign them, the employer is required to give the individual a copy of the report and a documentcalled “A Summary of Your Rights Under the Fair Credit Reporting Act(ftc.gov/bcp/edu/pubs/consumer/credit/cre35.pdf) (FTC Facts, 2010). Employers can communicatewith candidates in adverse action situations orally, in writing, or electronically and the notice mustinclude the following:

Table 4: FCRA Requirement

The name, address, and phone number of the company that supplied the credit report or background information; astatement that the company that supplied the information didn’t make the decision to take the adverse action and can’tgive you any specific reasons for it; and a notice of your right to dispute the accuracy or completeness of anyinformation in your report and to get an additional free report from the company that supplied the credit or otherbackground information if you ask for it within 60 days (FTC, 2010).

91


Failure to comply with FCRA requirements can lead to litigation initiated by the FTC, otherfederal agencies, states and individuals. Two companies recently settled complaints initiated byfired workers and rejected job applicants under FCRA. The companies, Quality Terminal Services,LLC and Rail Terminal Services, LLC agreed to b $53,000 and $24,000 in civil penalties (FTC,2009). In both cases, the companies conducted background checks that included credit reportssupplied by a consumer reporting agency. The employers allegedly took adverse action against theapplicants and employees and failed to provide the employees and applicants with pre-adverse actionnotices and adverse action notices as required by the FCRA (FTC, 2009).

EMERGING ISSUES

In recent years, a broad coalition of opposition to the use of credit checks as a part ofemployee selection has surfaced. The coalition includes the U.S. Equal Employment OpportunityCommission (EEOC), advocacy groups like the Lawyers’ Committee for Civil Rights Under theLaw, state legislatures, and the U.S. Congress (Maurer, 2010).

The EEOC’s involvement in this emerging issue is not new. The agency’s official positionposted on its web site is:

Table 5: EEOC’s Official Position

Title VII prohibits an employment practice that disproportionately screens out racial minorities, women, or anotherprotected group unless the practice is job related and consistent with business necessity. Thus, if an employer’s use ofcredit information disproportionately excludes African-American and Hispanic candidates, the practice would beunlawful unless the employer could establish that the practice is needed for it to operate safely or efficiently (EEOC,2010).

In the EEOC v. United Virginia Bank/Seaboard Nat’l case, the court determined that creditchecks were appropriate for certain positions, such as where the employee handles large amountsof cash even though the bank’s credit check policy disproportionately screened out African-American job applicants. The bank in this case was able to use a business necessity defense to thepractice because, according to the court, the bank had a business need to conduct pre-employmentcredit checks because employees handle large amounts of cash (EEOC v. United VirginiaBank/Seaboard Nat’l, 1977 WL 15340, 21 FEP Cases 1392 (E.D. Va. 1977). The position of theEEOC on the use of credit checks has been consistent with a similar opinion being expressed in aFebruary 14, 2005 opinion letter from EEOC Assistant Legal Counsel Dianna B. Johnston. In thisopinion letter, Johnston cited a 1974 case, United States v. City of Chicago, that held that a policedepartment could use financial information in background checks of applicants only if using theinformation did not have an “adverse impact” or is job related and consistent with businessnecessity(Johnston, 2005). Recently, in EEOC v. Freeman, the EEOC “initiated litigation against

92


a national employer alleging that the company engaged in a pattern or practice of unlawfuldiscrimination by refusing to hire a class of black, Hispanic, and male job applicants across theUnited States based upon sex, race and national origin due to the company’s use of credit checks,as well as criminal history checks” (Lewis, 2010).

In addition to initiating litigation, in February of 2007 the EEOC initiated the E-RACEInitiative “to bring a fresh, 21st century approach to combating racism, which remains the mostfrequent claim filed with the agency” (EEOC Press Release, 2007). In the press release announcingthe launching of the initiative, the EEOC made reference to studies that “some employers makeselection decisions based on names, arrest and conviction records, employment and personality tests,and credit scores—all of which may disparately impact people of color (EEOC Press Release, 2007).

In May of 2007, attorney Adam T. Klein made an important presentation at an EEOCmeeting in regard to the debate associated with employer use of credit checks of job applicants(Klein, 2007). Two of the most relevant questions Klein addressed for employers in his statementwere (1) Do Employee Credit Checks Have a Disparate Impact by Race? and (2) Are Credit ChecksJob-Related and Consistent With Business Necessity? (Klein, 2007). To the first question, Kleinprovided a great deal of empirical support for his conclusion that the answer is a resounding yes.Citing studies by Freddie Mac and others, Klein concluded that “there is a correlation between thequality of one’s credit record and one’s race” (Klein, 2007). Klein also concluded that “thecorrelation between credit record and race is only exacerbated by the fact that various credit“problems” correlate with race” (Klein, 2007). Problems identified by Klein include the fact thatcertain jobs (migratory work or low paying service jobs) are a minus for an individual’s creditrecord, bankruptcy filing, and lending discrimination in general which have all been shown tocontribute to poor credit ratings for African Americans (Klein, 2007).

Are credit checks job-related and consistent with business necessity? Klein presents a twopart analysis in responding to this question, concluding that under current legal standards firstdeveloped by the U.S. Supreme Court in Griggs v. Duke Power Co. and later in Albemarle PaperCo. v. Moody that the burden of proof that employers should be required to demonstrate to defendthe practice of employee credit checks is significant. According to Klein, to defend the use of creditchecks, “an employer would have to prove that it undertook a “meaningful study” that “validates”that credit record “bear{s} a demonstrable relationship to successful performance” (Klein, 2007).As to whether credit is job related and consistent with business necessity, Klein states “there is acomplete absence of evidence that employee credit checks are job-related at all, much less consistentwith business necessity, for any job” (Klein, 2007). He goes on to note, that in addition to the lackof evidence to support the validity of credit checks predicting job performance, he notes that “thereis substantial evidence that the credit records that employers check are based on factors substantiallyunrelated to any aspect of the performance of any job” (Klein, 2007).

Further, the reliability of the information supplied by credit reporting companies has beenfound to be inaccurate on numerous occasions (EPIC, 2008). The Electronic Privacy Information

93


Center (EPIC) cited a 1998 study done by US Public Interest Research Group (USPIRG) that foundthat 29% of credit reports contained serious inaccuracies including false judgments and falsedelinquency notices and that overall, 70% of reports “had some type of error” (EPIC, 2008). Whilemany organizations base their rationale for using credit checks on the notion of reducing orpreventing theft and embezzlement or to assess the trustworthiness of job candidates, Klein arguesthat “there is simply no support for the proposition that applicant (or incumbent employee) creditscore or history correlates to a heightened risk for theft” and that “African American applicants aremore likely to have bad credit” (Klein, 2007). Klein’s position is supported by Ben Arnoldy. In a2007 article, Arnoldy states that “so far, there’s a lack of data supporting a relationship between badcredit and theft by employees” (Arnoldy, 2007). Arnoldy cites a 2004 study published by twoEastern Kentucky professors that found a person’s credit history is not a good predictor of jobperformance or turnover (Arnoldy, 2007, & SIOP, 2004).

STATE LEGISLATIVE INITIATIVES

As of this writing, four states have enacted legislation limiting the use of credit checks byemployers in making selection decisions. On August 10, 2010, Illinois Governor Pat Quinn signedthe “Employee Credit Privacy Act” Public Act 096-1426. The effective date for the statute isJanuary 1, 2011. The meat of the statute is in section 10 of the act:

Table 6: Section 10 Illinois Public Act 096-1426

“Except as provided in this Section, an employer shall not do any of the following: Fail or refuse to hire or recruit,discharge, or otherwise discriminate against an individual with respect to employment, compensation, or a term,condition, or privilege of employment because of the individual’s credit history or credit report” (Illinois Public Act096-1426, 2010).

While the term employer in the statute is broadly construed, the act does provide exceptionsthat allows organizations like banks and credit unions or other similar business, companiesauthorized to engage in any kind of insurance or surety business pursuant to the Illinois InsuranceCode, state law enforcement agencies including the Department of Corrections, state or localgovernment agencies , or any entity that is defined as a debt collector under federal or State statuteto utilize credit check information when a satisfactory credit history is a bona fide occupationalrequirement for the job. Some of the specific circumstances that the act notes as bona fide includeduties of a position that include custody of or unsupervised access to cash or marketable assetsvalued at $2,500 or more, management positions that involve setting the direction or control of abusiness, positions that involve access to personal or confidential information, or the employee’sor applicant’s credit history is otherwise require by or exempt under federal or State law (Section10, Illinois Public Act 096-1426, 2010). In signing the bill, Governor Quinn was quoted as stating

94


that it will put a stop to the practice of denying jobs and promotions “based on information that isnot an indicator of a person’s character or ability to do a job well” (Fishman, 2010).

Three other states, Hawaii, Washington, and Oregon also have laws on the books similar toIllinois, and according to an Associated Press report, lawmakers in several states and the District ofColumbia advanced proposals in 2010 to prohibit employers from doing credit checks on most jobapplicants (Associated Press, 2010). The primary arguments proponents of these bills are advancinginclude potential discriminatory impact of employers’ use of credit check information in the currenteconomic climate. Representative Jack Franks, a sponsor of the Illinois statute stated in regard tothe need for the law, “it’s our neighbors, our families, our friends, we all know people who have losta job through no fault of their own and have struggled to pay bills” (Associated Press, 2010). StateSenator Diane Rosenbaum in Oregon noted that lawmakers in Oregon heard from a credit reportingagency official that “there was no indication of a link between [a} person’s credit history and theirjob performance or the likelihood of the person committing fraud” as “very compelling” in generating support for the Oregon law (Associated Press, 2010).

Table 7: State Legislatures Considering Legislation to Limit Employer Use of Credit Checks in 2010

California New Jersey

Connecticut New York

Georgia Ohio

Indiana Oklahoma

Louisiana Pennsylvania

Maryland South Carolina

Michigan Vermont

Missouri Wisconsin

Minnesota District of Columbia

Missouri

Source: National Conference of State Legislatures (NCSL), Use of Credit Information in Employment 2010Legislation, (NCSL, 2010).

While the state initiatives limiting the use of credit check information have generated a greatdeal of attention from employers, credit reporting agencies, and human resource professionals, noneof the statutes imposes an out-right ban on the use of credit checks and in fact, all provide for thegenerally accepted situations that courts and the EEOC have permitted so far.

95


U.S. CONGRESS INITIATIVES

Representative Steve Cohen (D-TN) proposed H.R. 3149 the Equal Employment for All Actin July of 2009. This piece of legislation was recently part of the discussion at a May meeting ofa U.S. House subcommittee on Financial Intuitions and Consumer Credit. The focus of the hearingwas on the “Use of Credit Information Beyond Lending: Issues and Reform Proposals (HR IssuesUpdate, 2010). Representative Cohen’s proposal would amend the Fair Credit Reporting Act tolimit the use of consumer credit checks for prospective and current employees for the purposes ofmaking adverse employment decisions (HR Issues Update, 2010). While Cohen’s proposal doesstate that it would “prohibit a current or prospective employer from using a consumer report or aninvestigative consumer report”, it does include in general terms most of the exceptions as toemployers and selected jobs that using credit check information would be permissible in makingselection decisions (H.R. 3149, 2010).

Table 8: Exceptions in H.R. 3149

[employment] (1) which requires a national security or Federal Deposit Insurance Corporation (FDIC) clearance; (2)with a state or local government agency which otherwise requires use of a consumer report; or (3) in a supervisory,managerial, professional, or executive position at a financial institution (H.R. 3149, 2010).

Amendments to the Fair Credit Reporting Act (FCRA) included in “The Wall Street Reformand Consumer Protection Act of 2009” signed into law by President Obama on July 21, 2010 alsomay impact employer use of credit information in making employment decisions. The impact ofH.R. 4173’s amendments to FCRA are not yet clear because the Federal Trade Commission (FTC)has not clarified language in the bill and updated its guidance and publications. At the FTC’s website, as of this writing, the primary guidance for employers was last updated and published in Mayof 2010. One initial report from an employment screening services practitioner is that the FCRAamendments in H.R. 4173 will require employers to provide “a copy of the credit report obtainedby the employer to their employee, and further explain to them what was wrong with the report thatresulted in the withdrawal of the employment offer” (Hassani, 2010). Also, prior to the Senate andHouse Conference version of the bill becoming law, Senator Diane Feinstein submitted SA 3795which would have amended the Senate version of the bill and FCRA. Senator Feinstein’samendment would have included a “General Prohibition” on the use of credit reports foremployment purposes or adverse actions, with limited exceptions. The amendment was ordered “tolie on the table” and would not be considered further by the Senate (SA 3795, 2010). This is furtherevidence that legislators at all levels of government are joining the bandwagon on this issue and thatemployers should be alert as new legislation progresses through the legislative process.

96


POLICY AND PRACTICE SUGGESTIONS FOR EMPLOYERS

Basic EEOC guidance with respect to the use of employment test and selection procedureshas been relatively consistent over time. With respect to the use of credit checks in the selectionprocess, current EEOC Chairman Stuart J. Ishimaru has been described as “a vocal critic ofemployee background checks, calling for the agency to issue guidelines within the next 12 to 18months on how to carry out background checks in a nondiscriminatory manner”(Maurer, 2010).Ishimaru has also stated “the issue of relying on credit checks in employment decisions remains anissue of concern for the EEOC and is likely to be raised more frequently in the coming years becauseof the ease of access to credit history information” (Lewis, 2010). Jackson Lewis believes that theEEOC “has made it clear” – “that an employer should be able to establish credit history informationis essential to the particular job in question, in addition to ensuring that it does not exclude anindividual from employment opportunity reflexively because of criminal or credit history” (Lewis,2010). Specific EEOC guidance includes the best practices suggestions in Table 9:

Table 9: Employer Best Practices for Testing and Selection

Employers should administer tests and other selection procedures without regard to race, color, national origin, sex,religion, age (40 or older), or disability.

Employers should ensure that employment tests and other selection procedures are properly validated for the positionsand purposes for which they are used. The test or selection procedure must be job-related and its results appropriatefor the employer’s purpose. While a test vendor’s documentation supporting the validity of a test may be helpful, theemployer is still responsible for ensuring that its tests are valid under UGESP.

If a selection procedure screens out a protected group, the employer should determine whether there is an equallyeffective alternative selection procedure that has less adverse impact and, if so, adopt the alternative procedure. Forexample, if the selection procedure is a test, the employer should determine whether another test would predict jobperformance but not disproportionately exclude the protected group.

To ensure that a test or selection procedure remains predictive of success in a job, employers should keep abreast ofchanges in job requirements and should update the test specifications or selection procedures accordingly.

Employers should ensure that tests and selection procedures are not adopted casually by managers who know littleabout these processes. A test or selection procedure can be an effective management tool, but no test or selectionprocedure should be implemented without an understanding of its effectiveness and limitations for the organization,its appropriateness for a specific job, and whether it can be appropriately administered and scored.

For further background on experiences and challenges encountered by employers, employees, and job seekers intesting, see the testimony from the Commission’s meeting on testing, located on the EEOC’s public web site at:http://www.eeoc.gov/abouteeoc/meetings/5-16-07/index.html (EEOC, 2008).

97


Bottom line for best practice for employers – “use credit information for employmentpurposes only if the credit information is job related – make individualized decisions – and notimpose a per se disqualification standard unless required by law to do so” (Lewis, 2010). NickFishman advises that in the current economic environment, use credit information as a “red flag”and “not to automatically exclude a candidate” (Fishman, 2010, A). If there are problems with acandidate’s credit report, given the current economic environment and the often suspect reliabilityof credit report information, “give the candidate a chance to explain” (Fishman, 2010, A).

In addition to following EEOC best practice advice, employers utilizing credit reports as partof employee selection, must be sure to comply with the provisions of FCRA. Non-compliance canget expensive as noted in recent FTC settlement agreements with companies that did not followproper notice procedures when taking adverse actions in employee selection. With recentamendments to FCRA, employers utilizing credit checks are advised to be alert for changes in policyand procedures forthcoming from the FTC.

Another important source of information for employers using credit checks will be EEOCcommunications. As noted above, the current Chairman is not a proponent of the use of creditchecks in employment decision making and has identified their use as an issue of concern for theEEOC. Employers must be alert for new guidelines with respect to the use of credit checks in thenear future. At the local level, employers in those states where legislation is being considered tolimit the use of credit information in employment, again, be alert for legislative action. The driveat the state level is in part being driven by the down economy and, with unemployment levels andeconomic growth still plaguing many cities and states, the prospect for new state laws remains alive.At the federal level, employers should also monitor the progress of Representative Cohen’s EqualEmployment for All Act as it moves through congress.

REFERENCES

Arnoldy, Ben (2007). The spread of the credit check as civil rights issue, Christian Science Monitor, January 18,downloaded 8/27/2010 from http://www.csmonitor.com/207/0118/p01s03-ussc.html.

Arnwine, Barbara R. (2006). Statement at EEOC meeting on race and color discrimination, April 19, 2006, downloaded11/6/2009 from http://www.eeoc.gov/abouteeoc/metings/4-19-06/foreman.html .

Associated Press (2010). Ban on credit checks on potential hires stalls in D.C., Md, downloaded 8/26/2010 fromhttp://www.washingtontimes.com/news/2010/jun/9/ban-credit-checks-..

Babcock, Pamela (2003). Spotting lies, HRMagazine, October Vol. 48, # 10, pp. 46-52.

Bates, Steve (2009). Caution urged on legislation to limit credit checks on job prospects, HR News, downloaded8/19/2010 from http://www.shrm.org/about/news/Pages/CreditChecksonJobProspects.aspx.

98


EEOC (2010). Title VII: Employer use of credit checks, downloaded 8/19/2010 fromhttp://www.eeoc.gov/eeoc/foia/letters/2010/titlevii-employer-creditck.html.

EEOC (2008). Employment tests and selection procedures, the U.S. equal employment opportunity commission,downloaded 8/19/2010 from http://www.eeoc.gov/policy/docs/factemployment_procedures.html.

EEOC Press Release (2007). EEOC Takes new approach to fighting racism and colorism in the 21st century workplace,downloaded 8/25/2010 from http://www.eeoc.gov/eeoc/newsroom/release/archive/2-28-07.html.

EEOC v. United Virginia Bank/Seaboard Nat’l, 1977 WL 15340, 21 FEP Cases 1392 (E.D. Va. 1977).

EPIC, (2008). The fair credit reporting act (FCRA) and the privacy of your credit report, EPIC – Electronic PrivacyInformation Center, downloaded 11/6/2010 from http://epic.org/privacy /fcra/ .

Fishman, Nick (2010). Illinois employers banned from credit reports (most of the time), downloaded 8/26/2010 fromhttp://www.employeescreen.com/iqblog/illinois-employers-can-no-longer-use-employment-credit-reports-most-of-the-time/ .

Fishman, Nick (2010, A). Credit reports: the value (and risk) to hr professionals, TLNT, downloaded 8/28/2010 fromhttp://www.tlnt.com/2010/07/21/credit-reports-the-value-and-risk-to-hr-professionals/ .

FTC Facts (2010). FTC Facts for consumers, Employment background checks and credit reports, downloaded 8/27/2010from http://www.ftc.gov.

FTC (2009). Two companies pay civil penalties to settle FTC charges; failed to give required notices to fired workersand rejected job applicants, downloaded 8/19/2010 from http://www.ftc.gov/opa/2009/08/qts.shtm .

Hassani, Reem (2010). New federal legislation to affect pre-employment credit checks, downloaded fromhttp://hire4integrity.premieress.com/blog-0/bid/40669/New-Federal-L...

HR Issues Update (2010). House examines use of credit information in employment, downloaded 8/19/2010 fromhttp://www.shrm.org/Advocacy/GovernmentAffairsNews/HRIssues..

H.R. 3149 (2010. Equal employment for all act, downloaded 8/27/2010 fromhttp://www.govtrack.us/congress/bill.xpd?bill=h111-3149&tab=summary.

H.R. 4173-737 (2010). The Wall Street reform and consumer protection act of 2009, public law No. 111-203,Section1100F, use of consumer reports downloaded 8/27/2010 from http://thomas.loc.gov/cgi-bin/query/z?c111:H.R.4173.

Illinois Public Act 096-1426 (2010). Employee Credit Privacy Act.

Johnston, Dianna (2005). Title VII credit reports – response to e-mail inquiry “is it okay for an employer to declineemployment due wholly or partly due to your credit report”, downloaded 11/6/2009 fromhttp://www.eeoc.gov/foia/letters/2005/titleii_credit _reports.html.

99


Klein, Adam T. (2007). Statement on May 16, 2007 meeting of the U.S. equal employment opportunity commission onemployment testing and screening, downloaded 8/19/2010 from http://www.eeoc.gov/eeoc/meetings/archive/5-16-07/klein.html.

Lewis, Jackson (2010). New Oregon credit check law warrants nationwide review of employer practices, downloaded7/9/2010 from http://www.shrm.org/LegalIssues/StateandLocalResources/Pages/NE...

Maurer, Roy (2010). Federal lawmakers, enforcers set sights on background screening, SHRM, downloaded 8/19/2010from http://www.shrm.org/LegalIssues/FederalResources/Pages/BackgroundScreening.aspx .

Minton-Eversole, Theresa (2008). Background screens even more crucial during economic slump, SHRM, downloaded8/19/2010 from http://www.shrm.org/hrdisciplines/staffingmanagement/Articles/Pages...

NCSL, (2010). Use of credit information in employment 2010 legislation, downloaded 8/27/2010 fromhttp://www.ncsl.org/default.aspx?tabid=19825.

SA 3795, (2010). Feinstein amendment SA 3795, Congressional Record, Senate Proceedings, page S3106, May 4,2010. Final action downloaded 8/27/2010 from http://thomas.loc.gov/cgi-bin/query/z?r111:S04MY0-0033 .

SHRM (2010). Background checking: conducting credit background checks, downloaded 8/19/2010 fromhttp://www.shrm.org/Research/SurveyFindings/Articles/Pages/BackgroundChecking.aspx .

SIOP (2004). Credit history not a good predictor of job performance or turnover, Society for Industrial andO r g a n i z a t i o n a l P s y c h o l o g y ( S I O P ) , 1 / 1 6 / 2 0 0 4 , d o w n l o a d e d f r o mhttp://.newswise.com/articles/view/502792?print-article .

100


101


DEVELOPING COLLABORATION SKILLS:A MIXED TEMPERAMENT APPROACH TO

TEAMWORK

David W. Nickels, University of North AlabamaJoan B. Parris, University of North Alabama

Carol H. Gossett, University of North AlabamaPaulette A. Alexander, University of North Alabama

ABSTRACT

Successful teamwork requires collaboration and the application of interpersonal andcommunication skills. Research shows that interpersonal and communications skills are highlydesired for information systems graduates by employers. Research also shows that there is valuein having students participate in teamwork projects while in school to develop more effectiveinterpersonal and communications skills and understand collaborative processes. Organizing teamsby mixing personality temperaments introduces an additional aspect of diversity providing a contextwhere students are challenged to strengthen interpersonal and group dynamic skills. This paperpresents an analysis of student experiences in using David Keirsey’s temperaments assessment(available at no cost online at www.keirsey.com) as the basis for team membership assignments inboth online and regular classroom settings. The feedback collected through student reflections onthe personality temperament assessment experience reveals that this self-discovery process is animportant step for students in developing better interpersonal and communication skills that theywill need to be successful in their careers.

INTRODUCTION

Information Systems (IS) literature contains extensive evidence that prospective employersof graduates from IS programs have high expectations for graduates’ interpersonal andcommunications skills (Abraham et al., 2006; Albin & Otto, 1987; Bell & Richter, 1987; Bullen etal, 2009; Watson, et al., 1990). College professors know that adding teamwork projects to students’coursework has the potential to provide them with opportunities to enhance those skills throughexperience in team environments with diverse members. This exploratory study focuses on thepersonality temperament aspect of diversity. The writers of this article have both investigated andintroduced the use of a personality temperament assessment method for building more diverse

102


student teams in both online distance learning classes and traditional face-to-face IS classes, witha goal of producing better-functioning teams and IS students who are better prepared tocommunicate and work well with fellow employees in the workplace after graduation. While otherdisciplines recognize and value teamwork skills (Mohamed & Lashine, 2003; Landrum & Harrold,2003; O’Brien & Deans, 1995), the present focus is on IS—the teaching discipline of the authors.This exploratory study within an IS environment is an expanded and updated version of "BuildingBetter Teams in both Online and Regular Classes by Mixing Personality Temperaments" previouslypublished in The Journal of Learning in Higher Education (2007).

BACKGROUND

Over the last two decades, multiple researchers have reported the finding thatcommunications skills are deemed critical for employers seeking new IS hires (Albin & Otto, 1987;Bell & Richter, 1987; Bullen et al., 2009; Watson, et al., 1990). Bailey & Stefaniak (2000), Doke& Williams (1999) and Van Slyke, Kittner & Cheney (1998) have written about the need forInformation Systems graduates to develop their interpersonal and communications skills in orderto meet industry’s needs. A graduate’s ability to think critically, write, and work effectively inteams is demanded by employers (Gruba & Al-Mahmood, 2004).

These communications skills have been placed at a comparable level of importance withtechnical skills by some authors. Verbick and Todd observe that “While a solid technological baseis desirable when hiring a computer lab consultant, communication and interpersonal skills are farmore crucial than programming expertise” (Verbick & Todd, 2003, p. 225). In terms of criticality,ranking communication skills higher than technical skills in systems development was also foundin a study by Ehie in 2002. He stated that “employers are looking for graduates with goodcommunications and people skills….Although technical-oriented skills were considered important,business-oriented skills were considered more important in hiring MIS graduates” (Ehie, 2002, p.154). In essence, mastering communication skills allows the students to communicate theirspecialized knowledge and collaborate with others to support the mission of the organization orbusiness at which they are employed (Baugh & Davis, 2008).

Evidence that IS academicians have widely recognized that need is represented by itsinclusion along with team skills in the recommended exit characteristics of IS graduates in the IS2002 Model Curriculum and Guidelines for Undergraduate Degree Programs in Information Systems(Gorgone et al., 2002). Despite the fact that communications skills in IS graduates have beenrecognized as a critical characteristic for new IS hires, those skills have been reported by prospectiveemployers of IS-related program graduates to be frequently lacking (Abraham et al., 2006). Inaddition, it has been observed that IS academics place a lower priority on the development ofcommunications skills than do IS professionals (Lee et al., 2002).

103


Confirming the critical importance of these skills, the Computing Accreditation Commissionof ABET, Inc. specifically addresses this issue in its accrediting guidelines for computing programs.Two of ten explicit IS program outcomes directly address the need for IS programs to enablestudents to achieve effective communication and teamwork skills by graduation. Specifically, thoseoutcomes are: “(d) An ability to function effectively on teams to accomplish a common goal” and“(f) An ability to communicate effectively with a range of audiences” (ABET, 2009).

Students who successfully participate in teamwork projects while in school are equipped tostrengthen key communication skills to help them not only to survive but to excel in their futurecareers. Crawford and Williams (2002) provide a clear delineation of the needed skills: “Gets alongwell with others, a team player, good communication skills, a focus on reality over theory,motivation.” It is likely that these skills can be acquired better through interactive team activitiesrather than passive, straight lecture formats. Hackbarth (1996) concluded that students’ learning theimportance of sharing and working collaboratively may very well be more important than learningprescribed subject matter.

STRATEGIES FOR ASSIGNMENT OF TEAM MEMBERS IN CLASSES

Teamwork is definitely an important part of a student’s training while in school. The presentresearch focuses on the question “how should class teams be formed for teamwork projects?” Foryears, a variety of methods have been used by teachers in traditional classes for team selections,including the following: (1) The students self-select their teams (sign up for the teams), (2) Theteacher assigns the team members based on who he/she thinks will be good leaders and goodfollowers, and (3) The teacher uses some randomization routine to form teams.

Based on many years of collective teaching experience, the authors of the present paper haveindependently concluded that flaws exist with each of the typical team-member assignment methods.Self-selecting by students for teams is not the best method for selecting team memberships becausestudents are most likely to want to be on teams with their friends and with people of similar interestsand work ethic, often resulting in one-dimensional teams with everyone thinking of the samesolutions, agreeing quickly with decisions because of friendships, or being frustrated by failure offriends to properly perform. Teachers assigning teams often do not have the knowledge ofleadership and other traits of all students in the class, rendering the approach of making carefulassignments ineffective. Randomization in creating teams is often viewed as more closelymimicking the workplace situation and potentially offers some opportunities for students toexperience workgroup diversity.

As an alternative to the standard team assignment models, these authors have used anapproach whereby students are assigned to teams based on diversity of personality types. Thatteams made up of a variety of personality types are often more powerful because the members havedifferent ways of looking at problems and their solutions has been confirmed in research.

104


Heterogeneity among team members in personality type has been found to be an importantdeterminant of team performance (Bradley & Hebert, 1997; Gorla & Lam, 2004; Neuman, Wagner& Christiansen, 1999; White, 1984). Specifically, this research has shown that teams composed ofmembers with diverse personality types are better performing than teams with members lacking thisdiversity. Neuman, Wagner, and Christiansen concluded that “a team will be more effective whenthe personalities of its members are diverse and each member contributes unique attributes to theteam” (p. 40).

ASSESSMENT OF TEMPERAMENTS OF CLASS MEMBERS

The strategy employed in this exploratory study involves having students complete apersonality temperament assessment prior to team formation. Stokes (2001, p. 24) suggests that“Identifying temperaments helps in understanding why individuals process and respond to the samesituations differently.” There are many good temperament studies and tests available for use, butthe Myers-Briggs Type Indicator (MBTI) and the Keirsey Temperament Sorter have been identifiedas the most popular instruments (Francis, Robbins, & Craig, 2007). The MBTI is widelyacknowledged as the most frequently administered personality type assessment instrument. Whilesome have questioned the validity of the MBTI, others have found strong support for its constructvalidity (Carlson, 1985; Thompson & Borello, 1986). Strong positive correlations betweenassessments of personality type by the MBTI and the Keirsey Temperament Sorter have also beenfound (Kelly & Jugovic, 2001), and these assessment methodologies have been identified asmeasuring the same constructs (Quinn, Lewis, & Fischer, 1992; Tucker & Gillespie, 1993). Inaddition, the Keirsey Temperament Sorter is available for student access at no cost online (Keirsey,2010). Therefore, the present authors have selected the Keirsey Temperament sorter as the tool forengaging students in an understanding of team dynamics, for studying personality types, and forassigning members to teams. Before beginning teamwork, students learn that:

“…people are different from each other, and that no amount of getting after themis going to change them. Nor is there any reason to change them, because thedifferences are probably good, not bad” (Keirsey, 2010).

Keirsey (2010) identifies four main temperament groups: Guardians, Artisans, Idealists, andRationals (Table 1). According to Keirsey (2010), “…it is important to understand that the fourtemperaments are not simply arbitrary collections of characteristics, but spring from an interactionof the two basic dimensions of human behavior: our communication and our action, our words andour deeds, or, simply, what we say and what we do.” The four temperaments have been furthersubdivided by Keirsey into “Character Types.” Keirsey identifies four types of Artisans, four types

105


of Guardians, four types of Rationals, and four types of Idealists as indicated in Table 1 (Keirsey,2010).

Keirsey’s Guardians are estimated to make up 40-45% of the world’s population. AllGardians, according to Keirsey (2010), share the following core characteristics:

! Guardians pride themselves on being dependable, helpful, and hard-working.! Guardians make loyal mates, responsible parents, and stabilizing leaders.! Guardians tend to be dutiful, cautious, humble, and focused on credentials and

traditions.! Guardians are concerned citizens who trust authority, join groups, seek security,

prize gratitude, and dream of meting out justice.

Keirsey’s Artisans make up 35%-40% of the population. All Artisans, according to Keirsey(2010), share the following core characteristics:

! Artisans tend to be fun-loving, optimistic, realistic, and focused on the here and now! Artisans pride themselves on being unconventional, bold, and spontaneous.! Artisans make playful mates, creative parents, and troubleshooting leaders. ! Artisans are excitable, trust their impulses, want to make a splash, seek stimulation,

prize freedom, and dream of mastering action skills.

Eight to ten percent of the population is estimated to be Idealists by Keirsey (2010). AllIdealists, according to Keirsey (2010), share the following core characteristics:

! Idealists are enthusiastic, they trust their intuition, yearn for romance, seek their trueself, prize meaningful relationships, and dream of attaining wisdom.

! Idealists pride themselves on being loving, kindhearted, and authentic. ! Idealists tend to be giving, trusting, spiritual, and they are focused on personal

journeys and human potentials.! Idealists make intense mates, nurturing parents, and inspirational leaders.

The smallest percentage of the population (5% to 7%) is Rational (Keirsey, 2010). AllRationals, according to Keirsey (2010), share the following core characteristics:

! Rationals tend to be pragmatic, skeptical, self-contained, and focused on problem-solving and systems analysis.

! Rationals pride themselves on being ingenious, independent, and strong willed. ! Rationals make reasonable mates, individualizing parents, and strategic leaders.

106


! Rationals are even-tempered, they trust logic, yearn for achievement, seekknowledge, prize technology, and dream of understanding how the world works.

Table 1: A Summary of Keirsey’s Four Temperament Categories and Subcategories (Keirsey, 2010)

Keirsey’s FourTemperaments:

Keirsey’s Four Subcategories WithinEach Temperament:

Characteristics of Keirsey’s Four Temperaments:

Guardian

Inspector Guardians:• Concrete in communicating and

cooperative in implementing goals• Can become highly skilled in logistics• They are often supervising and

inspecting or supplying and protecting

Protector

Provider

Supervisor

Artisan

Composer Artisans:• Concrete in communicating and

utilitarian in implementing goals• Can become highly skilled in tactical

variation• Usually promoting and operating or

displaying and composing

Crafter

Performer

Promoter

Idealist

Champion Idealists:• Abstract in communicating and

cooperative in implementing goals• Can become highly skilled in diplomatic

integration• Usually teaching and counseling or

conferring and tutoring

Counselor

Healer

Teacher

Rational

Architect Rationals:• Abstract in communicating and utilitarian

in implementing goals• Can become highly skilled in strategic

analysis• Involved in marshalling and planning

Fieldmarshal

Inventor

Mastermind

In the research design for this study, students first determine their individual personality bytaking the online Keirsey Temperament Sorter II (Keirsey, 2010). The students then compare theirpersonal results with the Keirsey Website dialog that describes their temperament types (Keirsey,2010). The results within the four main temperament categories are free, and if the students wantin-depth sub-temperament group information they can order and pay for that information. However,rather than asking students to pay for additional information, students in the classes are asked tostudy the four sub-temperament groups within their main temperament category and suggest whichgroup they think they fit into the best.

107


ASSIGNMENT OF STUDENTS TO TEAMS BASED ON THE TEMPERAMENT ASSESSMENT

To facilitate the learning process for students, following the online assessment each studentin the classes writes a short, one-page reflective paper on his or her assessment of the accuracy ofthe temperament identified in the results from the online questionnaire. In their papers, the studentsare asked to address ways in which the Keirsey Temperament results compare favorably orunfavorably with their view of their real temperaments. The papers are posted to an onlinediscussion board within the course’s learning management system (such as Blackboard or WebCT)for online distance learning classes, or they can be discussed in informal presentations in regularface-to-face classes.

Once the students have a better understanding of the characteristics of their owntemperaments, they are now ready to put their “gifts” to good use in teamwork exercises. Thewriters of this article have used teams for business case studies, for use in critiquing research projectplans, and for critiquing research surveys or questionnaires. There is ample evidence from theliterature on personality impacts on team performance to support the proposition that in teamperformance, the personality mix among team members does, indeed, matter (for example, seeBarrick, et al., 1998, and Harrison, et al., 2002). A student needing help in evaluating a researchsurvey that will be used in a research project will produce a stronger, more powerful surveyinstrument if it has been evaluated by others from different viewpoints and if that feedback isapplied.

In providing a means for organizing teams utilizing personality temperaments, the professormust first determine how many students in the class fall into the four main Keirsey temperamentcategories of Guardian, Artisan, Idealist, and Rational (Keirsey, 2010). Teams should have a varietyof personality temperaments and include all four temperaments if possible. However, sinceGuardians make up about 40-45% and Artisans make up 35-40% of the population (Keirsey, 2010),it is often necessary to have extra Guardians and/or Artisans on the team. For example, a typicalfive-member team might have two Guardian slots, one Artisan slot, one Idealist slot, and oneRational slot. Or another team may have one Guardian slot, two Artisan slots, and one slot each foran Idealist and a Rational.

Occasionally, depending on the temperament makeup of the class, it might be necessary tobreak one of the temperament categories down into the temperament sub-categories and have a five-member team such as this: one Guardian slot – subcategory of Supervisor, one Guardian slot –subcategory of Provider, one Guardian slot – subcategory of Protector, one Artisan slot, and oneIdealist OR Rational slot. As an alternative to the typical methods of team formation detailedpreviously, teachers may prefer to make their own team assignments based on the temperamentassessment results.

108


RESEARCH METHODOLOGY

The present research assesses student response to the use of a formal process of identifyingstudent personality type as a basis for team formation. Collectively, the writers of this article haveextensive experience with using mixed student personality temperaments as a method of teamselection in a variety of student learning settings and are driven by a belief that students need toexperience successes in teamwork situations. The mixed temperament approach to the team memberassignment process has been used in high school and in university undergraduate and graduateclasses. Of interest in the present research is the potential impact of the use of the KeirseyTemperament Sorter assessment process as the basis for assigning members to teams. The overallgoals of the temperament assessment include improved understanding of the behavior of self andothers in the team assignment environment and successful completion of the actual team assignment.This exploratory study analyzes reflective papers of 64 students in a convenience sample frommultiple classes during multiple semesters taught by two of the four authors of this paper. Thestudents include both graduate and undergraduate students in IS courses designed for students withIS concentrations, majors, or minors. All students in the subject classes were introduced to theKeirsey Temperament Sorter, completed the online questionnaire, received results, and performeda self analysis as described below. The reflective self analyses were the basis of an independentcontent analysis performed by three of the authors reviewing all of the responses and responding tothese questions:

1. Did the students agree with the results of the temperament sorter?

2. To what extent did the students agree with the results of the temperament sorter?

3. What were the key traits upon which the students based their agreement (ordisagreement), giving evidence of their buy-in?

To identify the answers to these questions, the narrative text provided by studentsparticipating in this exercise from multiple classes over multiple semesters was pooled into four files(one for each personality type) populated by student narratives based on the students’ assessedpersonality type. Next, a content analysis was performed on the individual text files by three of theseresearchers/authors to identify perceptions of accuracy and level of agreement (or disagreement).Content analysis is a commonly used and important qualitative research analytical technique (Cohen,Manion & Morrison, 2007; Krippendorff, 2004).

The researchers conducting the content analysis used color highlighting within the wordprocessing software to select positive statements of agreement with the temperament sorter resultsby the students. The researchers used a different color highlighting to select statements of

109


disagreement with the temperament sorter results. Another color highlighting was used to selectstudent comments about their key traits as they related to the temperament sorter results. Thecontent that was highlighted was then counted by each of the three authors and tallied by the fourmain Keirsey temperament categories of Guardian, Artisan, Idealist, or Rational. Then, the threeresearchers’ results were totaled and compared. With minimal differences found among the totalsof the three researchers, the averages of the three researchers’ results were then determined for eachcategory.

RESEARCH RESULTS

Based on the Keirsey temperament assessment, the distribution of personality types in thestudy sample as compared with the population as a whole is reported in Table 2. Differences in thedistribution might be attributed to possible bias introduced by the students’ review of thetemperament definitions prior to answering the questions on the instrument. In a self-reportinginstrument such as the Keirsey Temperament Sorter II, there is always the possibility of bias bystudents answering the questions in a manner that they think their friends or some mentor wouldlikely answer. The results in Table 2 show an unusually high number of students in the KeirseyIdealist and Rational categories compared to Keirsey’s population projections. It is important toremember that these students in the study were students in information systems and may have higheror lower concentration of traits in certain temperaments. These differences in distributionpercentages for the four temperaments are addressed in the Research Results section below.

Table 2. Distribution of Personality Types (Sample data and Keirsey, 2010)

Distribution of Personality Types

Personality Type Number in Study Sample Percent in Sample Percent of Population atLarge

Guardians 33 52% 40-45%

Artisans 9 14% 35-40%

Idealists 12 19% 8-10%

Rationals 10 16% 5-7%

TOTAL 64 100%

In relation to the question of the students’ perception of the accuracy of the temperamentsorter assessment of their personality types, the students in this study overwhelmingly agreed (92percent, as shown in Table 3) that the Keirsey assessment was accurate in describing theirpersonality temperaments. The content analysis revealed affirmative terms from the students suchas “I agree,” “a perfect match,” “is correct,” “I concur,” and “100% true.” Among the four Keirsey

110


assessment types, 94% of the assessed Guardians, 100% of the assessed Artisans, 92% of theassessed Idealists, and 80% of the assessed Rationals agreed with the results.

While no empirical attempt has yet been made to explain the differences between thetemperament types assessed in this study and Keirsey’s population projections, it is interesting tonote that the sample contains a high percentage of Rationals (16%), which is three times greater thanthe general population expectations of between 5-7%. Because of the nature of the work in the ISfield, it is likely that a higher number of Rationals might be attracted to the field. Many of thecharacteristics desired in IS professionals are those found in Keirsey’s descriptions of the Rationals.For example, Keirsey (2010) described Rationals as “…focused on problem-solving and systemsanalysis.” The focus of this study on IS students may also explain the lower than normal percentageof Artisan temperaments in the study. It is possible that very few of the terms used by Keirsey todescribe the Artisans would be used in describing most IS professionals, and students with Artisantemperaments might not be as attracted to the IS careers as other temperament groups.

Table 3. Distribution of Personality Types (Sample data)

Responses of Agreement Reported by Keirsey Assessment Personality Types

Personality Type Number in StudySample

Number Agreeing with theResult

Percent ExpressingAgreement

Guardians 33 31 94%

Artisans 9 9 100%

Idealists 12 11 92%

Rationals 10 8 80%

TOTAL 64 59 92%

The content analysis also revealed that the majority of the students did “buy in” to the self-discovery aspect, based on traits which they self-identified as confirmatory to the assessment results.Several common descriptive terms within each temperament category indicated that the studentswere benefiting from the self-discovery aspect of the assignment. For example, 67% of the studentsin the Rational category confirmed that they had strong characteristics of leadership and thrived onsetting objectives and accomplishing goals. One interesting pattern that was indicated by theresearchers’ content analysis coding involved only 33% of the Rational students, but it clearlyindicated a trend that some students in this category prefer to work alone. “Curiosity” and a lovefor “figuring things out” were indicated by 33% of the Rationals.

Of the Idealists in the study, 83% indicated that they enjoy helping others. Fifty percent ofthe Idealists mentioned that they were very ethical people who held themselves to strict standardsof personal integrity. There was a strong desire (75%) among the Idealists to be inspirational toothers and to encourage cooperation among others.

111


Ninety percent of the Artisans in the study confirmed that they enjoyed creating and doingthings with their hands. The hands-on work covered several areas such as art, music, and workingwith tools. Sixty percent of the Artisans described in a variety of terms that they would always dowhatever it takes to get the job done.

Of the Guardians in the study, 30% indicated that they were comfortable being “in charge,”“in control,” or “taking the lead” in group activities. Sixty-one percent of the Guardians discussedthe fact that they liked order and smooth operations. According to 30% of the Guardians, they liketo have fun but they know when to be serious and get their work done. A smaller percentage (24%)of the Guardians mentioned that they were “serious” about their jobs, school projects, andextracurricular activities work. An even smaller percentage (15%) considered themselves to be“reserved” and “quiet.” Thirty percent of the Guardians were not fond of change. Guardians tendto think of themselves as hard workers as indicated by 42% of those in the study.

In each of the assessed temperament categories, the students identified their own behaviorsand traits with remarkable insight. It was also interesting to note that they even demonstrated thecategory by their actions in writing the narratives. For example, the Rationals as a group had acommon trait that was clearly indicated in the content analysis coding for their descriptions of theirtemperament results: their analyses were shorter than the other categories and many simplyresponded as if to a very structured question. The content analysis reviewers observed that theGuardians, on the other hand, generally gave longer responses, noting more than three unique traitsor behaviors on average in support of their identification with the category. In addition, theGuardians elaborated far more extensively than the Rationals, including personal examples, sub-category identification, and often repeating the same trait in multiple ways. The Guardians alsofrequently included indications of how they have developed over time and how they might applythese findings to their own situations.

By contrast, the smaller percentile two categories—Idealists and Artisans—averaged about2 unique traits in each description. The unique traits expressed by those students were strikinglysimilar to the descriptions provided by Keirsey. Different between them, however, was the styleadopted to describe their traits and behaviors. For example, half of the Idealists began theirnarratives with a statement of what idealists are and the other half began with some form of “I aman Idealist.” Most of the assessed Idealists personalized with very descriptive words and phrases,expressing feelings and attitudes. Almost all of the Artisans began with a reference to their findingthat they fit into this particular category. The narrative responses for most of the Artisans appearedto reflect an attitude of “living in the here and now.” That is, they seemed intent on accepting thefinding and building an argument to support the assessed result, rather than analyzing reasons toagree or disagree.

As described in the discussion of results above, there is ample evidence that the studentsaddressed in this study perceived a high degree of accuracy in their assessed temperament types.

112


These researchers believe this finding lends strong support for continuing this methodology in futureefforts in building more effective student teams.

LIMITATIONS AND FUTURE RESEARCH

Data for this study is limited to reflections on a process made by students in IS classes forpurposes other than this research paper. As with any exploratory study, a number of extraneousvariables not controlled for in this study may have impacted the overall findings. These extraneousvariables include demographic factors of age, gender, ethnicity, academic level classification, andacademic major. The collection of demographic data in future studies would help control for theseitems. In addition, the results may have been impacted by differential conditions among multipleinstructors and multiple course sections. Finally, the meta-communication among the students, bothin person and via the electronic discussion board used to share student reflections on the process,may have produced a Hawthorne effect among the students who completed the KeirseyTemperament Sorter assignment after the majority of other students had completed their assignmentsand posted their reflective papers on the electronic discussion board.

The authors are also aware that personality type is only one of many dimensions of the teamproject experience. An additional dimension of interest is the learning style of the team members.Additional research is needed to determine the potential impact of learning styles on teamperformance. Further, interactions might be identified regarding the effects of learning styles andpersonality type.

Despite the aforementioned potential limitations, these authors believe that our initial beliefswere confirmed, at least on a preliminary basis, by the results of the study. To further investigatethe value of the proposed mixed personality strategy for team member assignment, future study willbe needed. Additional data is being collected in classes using this strategy to evaluate in detail thereactions of students at the end of a team project with diverse teams. Unsolicited anecdotal evidenceto date suggests that students enthusiastically embrace the concept after the project is completed.The next stage of research will involve an investigation of learning outcome assessment associatedwith specific teamwork learning goals. As a corollary, this work will support continuousimprovement and accreditation activities efforts within the academic department. Rubrics designedto assess learning outcomes will be used in the planned research as the basis for investigating severalpotential advantages to using a mix of student temperaments when building teams: (1) the studentknows why he/she is on the team, (2) the student knows he/she has a role to fill on the team and thathe/she is not just a generic member of the team, (3) the student knows that his/her opinion isimportant and that it is okay to think differently and have different ideas than all of the other teammembers, and (4) the student knows that the team has the potential to be a stronger team with avariety of personality temperaments represented.

113


CONCLUSION

From a professor’s perspective, one of the most rewarding parts of using temperaments asa means of building better teams for teamwork projects in business classes is watching the students“step up to the plate” and fully appreciate their roles on the team. That process can be observed inface-to-face classrooms as teams meet and work on projects, and it can also be observed indiscussion board activity within online distance learning classes. For example, one student with anassessed Guardian temperament displaying those characteristics wrote to teammates “It is dueNovember 7 - so I am taking on this Guardian role (control) and asking that you read the attachedand get your comments back to me by November 1.”

Table 4: Example of team dialog on a WebCT course discussion board at the completion of a team casestudy project. The bold print areas tell the story.

Author: MATTHEWDate: Monday, November 7 2:41pmOk Team,I have attached the final copy of the paper with the work cited page. Vanessa and I have looked over it one moretime and I think the paper is ready to be turned in. This was a great group to work with and I want to thank you for getting me all the information to me ontime. I had so much information that it made putting the paper together very easy. So thanks again for beingsuch a great group to work with.Matt

Author: AMANDA Date: Monday, November 7 5:33pmI agree Matt. I think we had a great group and I enjoyed working with all of you. - Mandy

Author: KERRY Date: Monday, November 7 8:51pmviva la group! it was my pleasure being on team 1. good luck with the paper,-kerry

Author: VANESSADate: Tuesday, November 8 8:21amThis has been a good group. I was wondering how it would go being on a team in an online class, but itworked out great! Thanks everyone!

It is clear that prospective employers of IS graduates continue to express a strong desire foreffective communication and interpersonal skills in prospective new hires. The challenge for ISeducators in meeting this demand is to provide maximal opportunities for students to acquirethose skills in preparation for their professional careers following graduation. There is confirmationin the research literature of the importance of personality type heterogeneity among team members

114


for high levels of performance. One strategy for enhancing student opportunities for communicationand interpersonal skills development through participation in project teams is presented anddiscussed in this paper.

The content analysis of the student reflections on their personality assessment experienceshows that, in the majority, students enrolled in the courses included in this study both agreed withthe accuracy of the assessment results and “bought in” to the process. It may be concluded, then, thatthe previously detailed methodology of forming student teams is an effective strategy. In addition,this strategy represents what these authors believe is a step toward strengthening IS programs tobetter equip students with the communication and interpersonal skills needed to succeed in teamenvironments in the workplace.

As a final observation, the desired experiences of team members are captured in theunsolicited dialog in Table 4, captured after a successful team project was completed. Thisrepresents the epitome of a successful project and a worthwhile goal for all team project experiences.

REFERENCES

ABET (2009). 2008-2009 criteria for accrediting computing programs – new criteria. Retrieved February 25, 2009from: http://www.abet.org/Linked%20Documents-UPDATE/Criteria%20and%20PP/C001%2009-10%20CAC%20Criteria%2012-01-08.pdf

Abraham, T., Beath, C., Bullen, C., Gallagher, K., Goles, T., Kaiser, K., & Simon, J. (2006). IT workforce trends:Implications for IS programs. Communications of the Association for Information Systems, 17, 1147-1170.

Albin, M. & Otto, R .W. (1987). The CIS curriculum: what employers want from CIS and general business majors. TheJournal of Computer Information Systems, 27, 15-19.

Bailey, J. L., & Stefaniak, G. (2000). Preparing the Information Technology Workforce for the New Millenium. ACMSIGCPR. Evanston, Illinois.

Barrick, M. R., Stewart, G. L., Neubert, M. J, & Mount, M. k. (1998). Relating member ability and personality to workteam processes and team effectiveness. Journal of Applied Psychology, 83(3), 377-391.

Baugh, J., Davis, G., & Turchek, J. (2008). Emphasizing Communication Skills in the Curriculum for the ComputerInformation Systems Grad. ISECON 2008, v25 (Phoenix, Arizona): 2553 (pp. 1-8). EDSIG.

Bell, J .D. & Richter, W.W. (1987). Needed: better communication from data processors. Personnel, 63, 20-26.

Bradley, J. & Hebert, F. (1997). The effect of personality type on team performance. Journal of ManagementDevelopment, 16, 337-353.

Bullen, C. V., Abraham, T., Gallagher, K., Simon, J., & Zweig, P. (2009). IT workforce trends: Implications forcurriculum and hiring. Communications of the Association for Information Systems, 24, 129-140.

115


Carlson, J. G. (1985). Recent assessments of the Myers-Briggs Type Indicator. Journal of Personality Assessment, 49,356-365.

Cohen, L., Manion, L. & Morrison, K. (2007). Research Methods in Education. London: Routledge.

Crawford, G., & Williams, R. (2002). Interactive teaching and course content: are we teaching the right things incolleges of business? Florence, AL: University of North Alabama. Retrieved May 14, 2007, from:http://www2.una.edu/gcrawford/mk360/ paper.htm

Doke, E., & Williams, S. (1999). Knowledge and skill requirements for Information Systems professionals: Anexploratory study . Journal of Information Systems Education, Sprint , 10-18.

Ehie, I. C. (2002). Developing a management information systems (MIS) curriculum: Perspectives from MISpractitioners. Journal of Education for Business , 151-159.

Francis, L. J., Robbins, M., & Craig, C. L. (2007). Two different operationalisations of psychological type: Comparingthe Myers-Briggs Type Indicator and the Keirsey Temperament Sorter. In R. A. Degregorio (Ed.), Newdevelopments in psychological testing (pp. 119-138). New York, NY: Nova Science Publishers.

Gorla, N. & Lam, Y. (2004). Who should work with whom? Building effective software project teams. Communicationsof the ACM, 47(6), 79-82.

Gorgone, J. T., Davis, G. B., Valacich, J. S., Topi, H., Feinstein, D. L. & Longenecker, H. E. (2002). IS 2002 modelcurriculum and guidelines for undergraduate degree programs in information systems. Communications of theAssociation for Information Systems, 11, 1-52.

Gruba, P., & Al-Mahmood, R. (2004). Strategies for communication skills development. ACE '04: Proceedings of thesixth conference on Australasian computing education - Volume 30.

Hackbarth, S. (1996). The educational technology handbook: a comprehensive guide—process and products forlearning. Englewood Cliffs, NJ: Educational Technology.

Harrison, D. A., Price, K. H., Gavin, J. H., & Florey, A. T (2002). Time, teams, and task performance: Changing effectsof surface- and deep-level diversity on group functioning. Academy of Management Journal, 45(5), 1029-1045.

Kelly, K. R. & Jugovic, H. (2001). Concurrent validity of the online version of the Keirsey Temperament Sorter II.Journal of Career Assessment, 9(1), 49-59.

Keirsey, D. (2010). Different drummers. Retrieved June 8, 2010 from http://keirsey.com/Drummers.htmlKeirsey, D.(2010). The four temperaments. Retrieved June 8, 2010 from: http://keirsey.com/personalityKeirsey, D.(2010). The keirsey temperament sorter II. Retrieved June 8, 2010 from:http://www.advisorteam.com/temperament_sorterKrippendorf, K. (2004). Content analysis: An introductionto its methodology. Thousand Oaks, CA: Sage Publications.

Landrum, R, E., & Harrold, R. (2003). Faculty forum: What employers want from psychology graduates. Teaching ofPsychology, 30(2), 131-153.

116


Lee, S., Koh, S., Yen, D. & Tang H. (2002). Perception gaps between IS academics and IS practitioners: An exploratorystudy. Information & Management, 40, 51-61.

Mohamed E. K. A, & Lashine, S. H. (2003). Accounting knowledge and skills and the challenges of a global businessenvironment. Managerial Finance, 29(7), 3-16.

Neuman, G., Wagner, S., & Christiansen, N. (1999). The relationship between work-team personality composition andthe work performance of teams. Group & Organization Management, 24, 28-45.

O’Brien, E. M., & Deans, K. R. (1995). The position of marketing education: A student versus employerperspective. Marketing Intelligence & Planning, 13(2), 47.

Quinn, M., Lewis, R., & Fischer, K. (1992). A cross correlation of the Myers-Briggs and Keirsey instruments. Journalof College Student Development, 33(3), 279-280.

Stokes, S. (2001). Temperament, learning styles, and demographic predictors of college student satisfaction in a digitallearning environment. (Doctoral dissertation, The University of Alabama, 2001). Dissertation AbstractsInternational, 62, 03, 983.

Tucker, I. & Gillespie B. (1993). Correlations among three measures of personality type. Perceptual and Motor Skills,77 (2), 650.

Van Slyke, C., Kittner, M., & Cheney, P. (1998). Skill requirements for entry-level IS graduates: A report from industry.Journal of Information Systems Education, Winter , 7-11.

Verbick, T., & Todd, K. (2003). Tech-NO-Nerds: why the best student computer lab consultants are often not from thecomputer science department. Proceedings of the 31st annual ACM SIGUCCS conference on User Services (pp.225-227). San Antonio, Texas: ACM.

Watson, H., Young, D., Miranda, S., Robichaux, B. & Seerley, R. (1990). Requisite Skills for new MIS Hires. DataBase, 21, 20-29.

White, K. (1984). MIS project teams: An investigation of cognitive style implications. MIS Quarterly, 8, 95-101.

117


PROACTIVE PERSONALITY:ORGANIZATION VS CAREER COMMITMENT

Leila Messara, Lebanese American UniversityGrace K. Dagher, Lebanese American University

ABSTRACT

Using a sample of 101 currently employed MBA students, the researchers assessed therelationship between proactive personality and the different types of commitment i.e. affective,continuance, normative and career. The results indicated that proactive personality is significantlypositively related to affective, continuance, and career commitment. The study suggests that ifemployees with proactive orientation are empowered to make decisions like an entrepreneur, thiswill enhance their commitment to both career and organization.

INTRODUCTION

In these turbulent times, the struggle for corporate survival has added pressure onorganizations to reconsider the role of individuals as key players and their capacity to revitalize theorganizations. Researchers have agreed that individuals’ personalities shape their behavior; we cantherefore conclude that individuals with certain traits could help organizations with their efforts toadapt. One of such traits is proactive personality. According to Bateman and Crant (1993, p.105)proactive individuals “scan for opportunities, show initiative, take action, and persevere until theyreach closure by bringing about change.” However, if such individuals are needed for change, canwe then count on their commitment to the organization, or, in other words, how committed are they?This prompted the researchers to investigate whether a relationship exists between proactivepersonality and the different types of commitment (Affective, continuous, normative, and career),especially since some organizations have identified proactive oriented behavior as a requirement intheir hiring process (Campbell, 2000).

PROACTIVE PERSONALITY

In recent years, personality traits have become both popular and an accepted means forexplaining individuals’ behavior, i.e. actions, manners, targets, and purposes (Llewellyn and Wilson,2003). This helps identify the reasons for individuals’ different reactions to similar situations(Cooper, 1998). Ryckman defined Personality as the “dynamic and organized set of characteristics

118


of a person that uniquely influences his/her cognitions, motivations, and behaviors” (1982, p. 5).This concept represents behavioral and cognitive prototypes that have been proven stable throughtime and in different settings (Cattell, 1964). According to Bateman and Crant (1993) past researchhave considered proactive personality as a stable individual disposition.

Individuals with proactive oriented personality are somewhat unconstrained by situationalfactors and affect environmental change (Bateman and Grant, 1993).They recognize opportunities,show initiative, take action, and persist until meaningful positive change occur in their environmentregardless of obstacles (Seibert, Kraimer, & Grant, 2001). Proactive individuals tend to set highstandards and make use of all available resources toward achieving them (Grant, 1996). Crantdefined proactive behavior as “taking initiative in improving current circumstances or creating newones; it involves challenging the status quo rather than passively adapting to present conditions”(2000, p. 436). On the other hand, less proactive or reactive individuals tend to be passive, showlittle initiative, and are likely to adapt to situations rather than change their circumstances.

Past research has positively associated proactive behavior with entrepreneurship (e.g.,Becherer & Maurer, 1999; Grant, 1996), individual and job performance (e.g. Grant, 1996; Ashfordand Northcraft, 1992), career success (e.g. Seibert et al., 1999), and leadership (e.g. Crant andBateman, 2000; Deluga, 1998). Although most of these studies indicated positive outcomes withproactive personality, nevertheless, a study by Bolino and Turnley (2005) linked it with job stress,work family conflict, and job overload which are negative outcomes.

Commitment

The definition of commitment in the literature seems inconsistent. However, in this researchwe define organizational commitment as the employee’s goal to remain with the organization(Meyer & Allen, 1997). Organizational commitment has been positively linked to hours devoted towork, morale, absenteeism, intent to quit, performance of the organization (Silverthome, & Hung,2006; Fernanado J. et al., 2005; Riketta, Chen, 2002) and the satisfaction of employees (Gallie andWhite 1993). Previous literature also indicated that organizational commitment and job performancehave a direct relationship (Chan, D. 2006, Fernando J. et al., 2005, Vandenberghe, C. et al., 2004,Bishop J. W. et al., 2000), is positively related to the need for achievement (Lee, 1971;Patchen,1970), and job challenge (Buchman, 1974). Hence, maintaining highly committed personnelshould be a priority in the minds of those responsible.

Meyer & Allen (1997) distinguished a three-component model of commitment anddeveloped a scale to measure them which generally holds up across cultures (Sulimand, & ILes,2000; Ko, Price, & Mueller, 1997). These are: (a) Affective commitment which involves theidentification, involvement and emotional attachment with the organization leading to the sentimentof wanting to continue employment in the organization; (b) Continuance commitment which standsfor profit coupled with continued participation on the one hand, and the cost associated with leaving

119


on the other hand (Kanter, 1968) creating a feeling of needing to continue employment; and (c)Normative commitment represents a feeling of obligation towards the organization i.e. a personought to continue employment. Together, these components make up an employee’s ‘commitmentprofile’

Organization and job commitment have received a great deal of attention from researchers.However, few studies have examined career commitment in general, and in a non-western contextin particular. Researchers have indicated the distinctiveness of career, organization and workcommitment constructs; nevertheless, they have also indicated a correlation between them (Gouletand Singh, 2002; Morrow, 1983 &1993; Muller, Wallace, & Price, 1992; Wiener & Vardi, 1980).Career commitment reflects individual’s commitment to a specific work and could relate to workoutcomes (Ballout, 2009). According to Noordin, Williams and Zimmer (2002), Career commitmentis reflected by the individual’s identification with the career more than with the organizationmembership.

Career Commitment is the planned choice of a line of work and the belief that loyalty in thischoice will surpass a particular job or organizational context (Morrow, 1993). According toColarelli and Bishop (1990), career commitment measures not only the individual’s identificationof personal career goals and attachment to these goals but also measures the involvement in thesegoals. In this study, we adopt Blau’s (1985, p. 278) definition of career commitment as “one’sattitude toward one’s profession or vocation”.

HYPOTHESES

Proactive Personality and Commitment

An individual’s predisposition is basically important in understanding his/her tendencytoward certain action. A growing body of literature shows that a relationship exists amongpersonality variables and attitude or behavior at work (Roberts and Hogan, 2001), and commitmentto the organization is attitudinal and behavioral in nature.

Attitudinally, individuals identify with the organization and are committed to remain in orderto pursue goals (Porter et al., (1974); while behaviorally, individuals are bound to the organizationthrough diverse interest such as, seniority, pension, etc. (Becker, 1960).

Meyer and Allen (1997) argue that affective commitment is positively related to individuals’willingness to commit extra effort to their work; this is the kind of commitment that can be expectedto be related to proactive orientation. We propose:

H1: Proactive personality is positively related to affective commitment

120


On the other hand, normatively committed employees attach themselves to the organizationsolely because they believe it is the right way to behave. It is based on the individual’s personalobligation to act in a way to meet organizational goals and interests (Allen and Meyer, 1990;Wiener, 1982). Proactive individuals take action to influence positive change in their environment.They possess entrepreneurship initiatives which make them more likely to quit the organization tostart their own businesses (Becherer and Maurer, 1999). Thus, we can predict that:

H2: Proactive personality is not related to normative commitment

A longitudinal study by Seibert, Grant and Kraimer (1999) found that proactive orientationis positively associated with innovation, self reported objective (salary and promotion), andsubjective career satisfaction. Together, self reported objective (which are indicators of perceivedcost/benefit, referred to by Meyer and Allen as continuance commitment) and subjective careersatisfaction indicate career success. We can thus propose the following Hypotheses:

H3: Proactive personality is positively related to continuance commitmentH4: Proactive personality is positively related to career commitment

METHODOLOGY

Instruments

In this study, the researchers focused on understanding the relationship between proactivepersonality and the different types of commitment i.e. affective, continuance, normative and career.A five parts questionnaire was developed to include items to measure the three dimensions oforganizational commitment, career commitment and proactive personality. To measureorganizational commitment, the Component-model of Commitment developed by Meyer & Allen(1997) was used in this research for it was specifically designed to measure the three types ofcommitment to the organization; i.e. affective, continuance, and normative commitments. It is amultidimensional scale that is widely used and intensively tested (e.g. Culpepper, 2000; Jaros,1997). It includes 24 items, 8 items per commitment type. Sample items: affective commitment“This organization has a great deal of personal meaning to me”; continuance commitment “Rightnow, staying with my organization is a matter of necessity as much as a desire”; normativecommitment “Jumping from organization to organization does not seem at all unethical to me”. To measure career commitment the researchers used the scale developed by Blau (1985) consistingof eight items. Sample item “I definitely want a career for myself in this industry”.

As for proactive personality, we used a 10 item scale developed by Seibert et al., (1999).Sample item “I am always looking for better ways to do things”. All items were measured using a

121


7-point Likert scale ranging from 1 as strongly agree to 7 as strongly disagree. In addition to thescale’s items we included demographic questions as well.

Sample

The sample was composed of employed MBA students attending an American Universityin Lebanon. The researchers distributed 140 questionnaires but only 101 completed the surveyforming a 72% response rate. Of the sample, 53 were male and 48 female. The majority of therespondents were single and their ages were between 20 and 30 years old. As to their currentposition, 42 of the respondents have a supervisor position and 24 have a middle managementposition.

Hypothesis Testing Results

In accordance with previous results on the different types of commitment the reliabilitycoefficient scores for the different types of commitment were consistent with the previous studieson commitment. Table 1 provides the means, standard deviation and coefficient alpha for thevariables in this study.

Table 1: Mean, Standard Deviation, and Reliability coefficient for the different types of Commitment

Variables Mean SD 1 2 3 4 5

1. Proactive Personality 2.31 0.87 -0.88

2. Career comm 3.54 0.9 .356** -0.71

3. Continuance comm. 3.68 0.8 .326** .212** -0.75

4. Affective comm. 3.07 0.96 .489** .655** .214** -0.74

5. Normative comm 3.64 0.62 .206* .543** .207* .439** -0.73

* Correlation is significant at the .05 level ** Correlation is significant at the .01 level

In order to examine the predicted relationship between proactive personality andcommitment i.e. affective, continuance, normative, and career, a Pearson correlation analysis wasused. Data were analyzed using SPSS 17 statistical package. Hypothesis 1 predicted that proactivepersonality would be positively related to affective commitment. A significant positive associationof r=.509(p<.01) was found between proactive personality and affective commitment; thussupporting the predicted relationship. Hypothesis 2 predicted that proactive personality would haveno relation with normative commitment. A significant relation was found; thus H2 was notsupported. Hypothesis 3 predicted that proactive personality would be positively related to

122


continuance commitment. A significant positive association of r=.364 (p<.01) was found betweenproactive personality and continuance commitment; thus supporting the predicted relationship.Hypothesis 4 predicted that proactive personality would be positively related to career commitment.A significant positive association of r=.356 (p<.01) was found between proactive personality andcareer commitment; thus supporting the predicted relationship. The results of the correlation analysisare provided below in Table 2.

Table 2: Correlation Results between Proactive personality and commitment (N=101)

Career Continuance Affective Normative

Proactive Personality .356** .326** .489** .206*

*Correlation is significant at the 0.05 level. **Correlation is significant at the 0.01 level.

To further examine the proposed relations and to provide a complete understanding of theserelations regression analysis was conducted. The regression results provided additional support tothe results of the Pearson correlation. The results indicated that proactive personality significantlypredicted the four types of commitment. Table 3 provides a detailed description of the regressionresults.

Table 3: Regression analysis results

Hypotheses B β SE R² T Sig

H1: DV: Affective Commitment P- Proactive Personality

.564*** .489*** 0.1 0.23 5.612 0

H2: DV: Normative Commitment P- Proactive Personality

.133* .179* 0.07 0.02 1.819 0.07

H3: DV: Continuance Commitment P- Proactive Personality

.282*** .303*** 0.08 0.08 3.174 0

H4: DV: Career Commitment P- Proactive Personality

.341*** .324*** 0.148 0.1 3.429 0

*p<.1, **p<.05, ***p<.01

DISCUSSION AND CONCLUSION

A sample of 101 currently employed MBA students from an American university in Lebanonwas surveyed in order to assess the relationship between proactive personality and affective,continuous, normative and career commitment. Our results show that proactive personality issignificantly positively related to affective and continuance commitment. This could be the result

123


of our sample characteristic i.e. employed MBA students. In general, MBA students are young andas such might be either newly employed or has not been with the organization for a long time. Thisis in line with previous literature that found a positive relation between proactive personality andnewcomer adaptation (Chan and Schmitt, 2000).

Our results also indicate a positive relation between proactive personality and careercommitment. This is also in line with previous research that linked proactive orientation withentrepreneurship (e.g., Becherer & Maurer, 1999; Grant, 1996). In addition, although our resultsshow a significant relationship between proactive personality and normative commitment,nevertheless, the r square value is 2% indicating a very weak relation between these variables. In conclusion, the widely debated relationship between career commitment and organizationalcommitment remains controversial. While Aranya and Ferris (1984) suggest commitment to onlyone, Norris and Niebuhr (1983) and Bartol (1979) conversely see high commitment to both.Gouldner (1957), however, claims incompatibility.

As stipulated in our research, provided the empirical evidence that proactive personality waspositively related to affective, continuance, and career commitment, and since previous literaturehas positively related proactive orientation to entrepreneurship, we can therefore conclude: ifemployees with proactive orientation are empowered to make decisions like an entrepreneur, thiswill enhance their commitment to both career and organization. According to Chatman et al., (1998),individuals whose values are congruent with the operating values of the organization are more likelyto be committed to the organization.

LIMITATION AND FUTURE RESEARCH

Our results extend the literature on proactive personality and commitment. However, nostudy is without limitations. The first limitation is that data was collected from a single source thusthe study can not be generalized. It is recommended that future research should examine whetherour finding can be generalized beyond employed MBA students and other cultures. The secondlimitation is the demographic characteristic with regard to the age and the education level of ourrespondents, the majority of our sample was between 20-30 years old and pursuing their mastersdegree. Future research should collect data from different groups. The third limitation is the self-reported data. Future research should collect data from both employees and supervisors to have abetter understanding of the employees’ commitment to career and organization and how these arelinked to proactive personality. Further studies could also examine variables that might mediate therelationship between proactive personality and career commitment such as motivation and jobsatisfaction. In addition, examining the moderating effect of gender, age, and experience on therelation between proactive personality and the commitment types may provide a betterunderstanding of this type of relationships.

124


REFERENCES

Allen, N. J. & Meyer, J. P. (1990). The measurement and antecedents of affective, continuance and normativecommitment to the organization. Journal of Occupational Psychology, 63, 1-18.

Aranya, N. and Ferris, K.R. (1984). A reexamination of Accountants’ Organizational – Professional Conflict, AccountingReview, 59, 1-14.

Ashford, S.J., & Northcraft, G.B. (1992). Conveying more (or less) than we realize: The role of impression managementin feedback-seeking. Organizational Behavior and Human Decision Processes, 53, 310.–334.

Ballout, H.I. (2009). Career commitment and career success: moderating role of self-efficacy, Career DevelopmentInternational, 14 (7), 655-670.

Bartol, K.M. (1979). Professionalism as a Predictor of Organizational Commitment, Role Stress, and Turnover: AMultidimensional Approach, Academy of Management Journal, 22, 815-821.

Bateman, Thomas S., & J. Michael Crant (1993). "The Proactive Component of Organizational Behavior: A Measureand Correlates," Journal of Organizational Behavior, 14, 103-118.

Becherer,R.C.,& Maurer, J.G. (1999). The proactive personality disposition and entrepreneurial behavior among smallcompany presidents, Journal of Small Business Management, 28-36.

Becker, H. S. (I960). Notes on the concept of commitment. American Journal of Sociology, 66, 32-40.

Bishop, J.W., Scott, K.D. & Burroughs, S.M. (2000). Support, commitment and employee outcomes in a teamenvironment. Journal of Management, 26(6), 1113-1132.

Blau, G.J. (1985). The measurement and prediction of career commitment. Journal of Occupational Psychology, 58, 277-288.

Bolino,M.C.,& Turnley,W.H. (2005). The personal cost of citizenship behavior: the relationship between individualinitiative and role-overload, job stress, and work-family conflict. Journal of Applied Psychology, 90(4), 740-748.

Buchman, B. (1974). Building organizational commitment: the socialization of managers in work organizations,Administrative Science Quarterly, 19, 533-546.

Campbell, D.J. (2000). The proactive employee: Managing workplace initiative. Academy of Management Executive,14(3), 52.–66.

Cattell, R. B. (1964). Objective personality tests: A reply to Dr. Eysenck. Occupational Psychology, 38, 69-86.

Chan, D., & Schmitt, N. (2000). Interindividual differences in intraindividual changes in proactivity duringorganizational entry: A latent growth modeling approach to understanding newcomer adaptation. Journal ofApplied Psychology, 85, 190-210.

125


Chatman, J.A., Polzer, J.T., Barsade, S.G., Neale, M.A. (1998), Being different yet feeling similar: the influence ofdemographic composition and organizational culture on work process and outcomes, Administrative ScienceQuarterly, 43, 749-80.

Chan. D (2006). Interaction Effects of Situational Judgment Effectiveness and Proactive Personality on WorkPerceptions and Work Outcomes. Journal of Applied Psychology, 91(2), 475-481.

Chen, J., C. Silverthome & J. Hung (2006). Organization Communication, Job Stress, Organizational Commitment, andJob Performance of Accounting Professionals in Taiwan and America. Leadership & OrganizationDevelopment Journal, 27 (4), 242-249.

Colarelli, S.M. and Bishop, R.C. (1990). Career commitment: Functions, correlates, and management, Group &Organization Studies, 15, 158-176.

Cooper, H. (1998). The happy personality: A meta-analysis of 137 personality traits and subjective well-being.Psychological Bulletin, 124, 197–229.

Crant, J. M. (2000). Proactive behavior in organizations. Journal of Management, 26, 435-462.

Crant, M. J., & Bateman, T. S. (2000). Charismatic leadership viewed from above: The impact of proactive personality.Journal of Organizational Behavior, 21, 63-75.

Culpepper, R. A. (2000). A Test of Revised Scales for the Meyer and Allen (1991) Three- Component CommitmentConstruct, Educational and Psychological Measurement, 60 (4), 604-616.

Deluga, R. (1998). American presidential proactivity, charismatic leadership and rated performance. LeadershipQuarterly, 9, 265.–291.

Fernando J., Jay P. M., and Greg W. M. (2005). A meta-analysis of the relationship between organizational commitmentand salesperson job performance: 25 years of research, Journal of Business Research, 58,705-714.

Gallie, D., White, M. (1993). Employee Commitment and the Skills Revolution: First Findings from the Employmentin Britain Survey. London: Policy Studies Institute.

Gouldner, A.W. (1957), Cosmopolitans and locals: toward an analysis of latent social roles, Administrative ScienceQuarterly, 2, 281-306.

Goulet, L.R. & Singh, P. (2002). Career commitment: A Reexamination and an extension, Journal of VocationalBehavior, 61, 73-91.

Grant, R.M. (1996). Toward a knowledge-based view of the firm, Strategic Management Journal, 17,109-22.

Jaros, S. J. (1997). An Assessment of Meyer and Allen’s (1991) Three- Component Model of Organizational andTurnover Intentions, Journal of Vocational Behavior, 51 (3), 319-337.

126


Kanter, R. M. (1968). Commitment and social organization: a study of commitment mechanisms in Utopiancommunities, American Sociological Review, 33, 499-517.

Ko, J., J. L.Price & C. W. Mueller (1997). Assessment of Meyer and Allen’s Three Component Model of OrganizationalCommitment in South Korea, Journal of Applied Psychology, 82 (6), 961-973.

Lee, Sang M. (1971). An empirical analysis of organizational identification, Academy of Management Journal, 14, 213-226.

Llewellyn, D and Wilson (2003). The controversial role of personality traits in entrepreneurial psychology, K. Education+ Training, 45 (6).

Meyer, J. P. & Allen, N.J. (1997). Commitment in the Workplace: Theory, Research, and Application. Sage Publications,Thousand Oaks.

Morrow,P.C. (1993). The Theory and Measurement of Work Commitment, Jai Press Inc., 157-188.

Morrow, P.C. (1983) Concept redundancy in organizational research: The case of work commitment, Academy ofManagement Review, 8(3), 486-500.

Muller, C. W., Wallace, J.E., & Price, J.L. (1992). Employee Commitment: Resolving some issues. Work andoccupations, 19(3), 211-236.

Noordin, F., Williams, T., & Zimmer, C. (2002). Career commitment in collectivist and individualist cultures: acomparative study, International Journal of Human Resource Management, 13(1), 35-54.

Norris, D.R.and Niebuh, R.E. (1983). Professionalism, Organizational Commitment and Job Satisfaction in AccountingOrganization, Accounting, Organizations, and Society, 9, 49-59.

Patchen, M (1970). Participation, Achievement, and Involvement on the Job. Englewood Cliffs, N.J.: Prentice Hall.

Porter,L.W.,Steers, R.M.,Mowday,R.T.,& Boulian,P.V. (1974). Organizational commitment, job satisfaction, andturnover among psychiatric technicians, Journal Of Applied Psychology, 59,603-609.

Riketta, M. (2002).Attitudinal Organizational Commitment and Job Performance: A Meta-Analysis, Journal ofOrganizational Behavior, 23(3), 257-266.

Roberts, B. W., & Hogan, R. (Eds.). (2001). Personality psychology in the workplace.

Ryckman, R. (1982). Theories of Personality, (2nd ed.). Brooks-Cole, Monterey, CA.

Salgado, J. F. (1997). The five factor model of personality and job performance in the European Community. Journalof Applied Psychology, 80, 607-620.

Seibert,S.E., Crant, J.M.,& Kraimer, M.L. (1999). Proactive personality and career success, Journal of AppliedResearch, 416-427.

127


Seibert,S.E.,Kraimer, M.L. & Crant,J.M.(2001).What do proactive people do? A longitudinal model linking proactivepersonality and career success, Personnel Psychology, 850.

Sulimand,A.M. & P.A . Iles (2000).The Multi-Dimensional Nature of Organizational Commitment in a non-WesternContext, Journal of Management Development. 9(1), 71-82.

Vandenberghe, C. , Bentein, K. & Stinglhamber, F. (2004), Affective commitment to the organization, supervisor, andwork group: Antecedents and outcomes. Journal of Vocational Behavior, 64, 47-71.

Wiener, Y., & Vardi, Y. (1980). Relationships between job, organization, career commitments and work outcomes: Anintegrative approach, Organizational Behavior and Human Performance, 26, 81-96.

Wiener,Y.(1982). Commitment in organizations: A Normative View, Academy of Management Review, 7 (3), 418-428.

128


Allied Academies

invites you to check our website at

www.alliedacademies.orgfor information concerning

conferences and submission instructions